TLDR;

• The new Workday Flex Credits and Platform Entitlement Policy took effect May 30, 2026.

• The grace period everyone is misreading: from May 30, 2026 through January 31, 2027, policy signers pay no overage charges on Application APIs. That is the whole scope. It does not waive your agent credit burn.

• The meter only works if you finished the setup almost nobody finished. Sign the UMSA. Stand up your Agent System of Record. Switch on the Platform Consumption Console.

• Signing the policy quietly flips on Sana Core in Production. That includes a new native auth login experience, pushed outside the normal rollout schedule.

• The rate card is not a price list. It is a map. Some agents are cheap enough to explore all year. One payroll configuration can vaporize your entire complimentary allotment in a single run.

Credit where it is owed. I would not be writing this piece, and frankly I might not be writing at all, if Jess had not started pulling this thread (and others) in public first (Jess is the absolute boss on working an Agile-led product management Workday team. Read her stuff. It was and still is a major reason why I believe in practitioner-led Workday content). Jess is an HRIS Manager in Minneapolis. A practitioner. Not an analyst, not a vendor voice. What follows builds on a public, on-the-record conversation where she walked through negotiating her own organization’s Workday agreement out loud. That is the rarest kind of source in this space. Not a doc interpretation. A practitioner narrating the actual commercial process while she lived it. The framing here builds on her notes. The modeling is mine.

Let’s go.

1. The gate nobody mentioned

You can have a Flex Credits balance sitting in your tenant right now and be unable to spend a single one of it on a Workday-built agent. Because there is a separate legal unlock, and it is not part of standard order form execution.

It is the Universal Main Subscription Agreement. The UMSA.

Per Jess, this is the thing that quietly replaced the old separate Flex Credit application path. No more side door. The UMSA is the front door, and it covers the whole agent stack: the Agent System of Record, the agent gateway, the Flowise builder, all of it.

One clarification, because this is exactly where people get the wrong idea. The UMSA unlocks the use of your entitlements. It does not hand you more Flex Credits. If you want more, or you run past what you have, you sign a Flex Credit Order Form to buy them. And Workday is watching. It monitors your consumption through the PCC and will flag you as you approach your entitlement levels, with a nudge to start looking at purchasing credits before you run out. The gate lets you in. It does not refill the tank.

Here is the asymmetry that catches people. UMSA is required for non-production and production tenants for Workday-built agents. It is not required for non-production tenants for self-built agents. So your sandbox tinkering on a custom agent does not trip the requirement. The moment you take a Workday-native agent toward Production, it does.

Check your own status in about thirty seconds. On Workday Community, open your profile, click your organization name under your title, and read the Subscription Service Agreement value. If it says MSA, you have not opted in and you are gated. If it says UMSA, you are clear and the features are automatically available.

Most practitioners have never looked at that field. Go look.

2. The setup nobody finished

Signing the UMSA opens the door. It does not walk you through it. There are two configuration jobs waiting on the other side, and neither turns on by itself.

Job one: the Agent System of Record.

The ASOR is the control center for your digital agent workforce. Discover agents, register them, manage them, audit them. It is where governance actually lives. And standing it up is a manual security exercise across six domains.

Five live in the new Agent System of Record functional area:

• Agent Compliance

• Agent Management Hub

• Manage: Agents

• Reports: Agent Reporting

• Setup: Agents

One more lives in the System functional area:

• Reports: AI Agent Security

You enable the functional area, activate pending security changes, enable each domain security policy, grant View and Modify to your security groups, then activate again. One caveat that will bite you: today you can only configure unconstrained groups on ASOR security policies. And there is no migration tool. You repeat this in every tenant, by hand. Sandbox, implementation, Production. Each one.

Configure an agent and Workday auto-generates an Agent System User account for it. That ASU is how the agent authenticates. External agents, partner-built or your own, route through the new Agent Gateway on regional endpoints, OAuth 2 only. If you have integrations pointed at old endpoints, that is a migration conversation, not a checkbox.

Job two: the Platform Consumption Console.

The PCC is your meter. It is also off by default. To turn it on, a security admin creates a security policy on the Management Dashboard: Platform Consumption Console domain in the System functional area, adds the relevant security groups, grants View, and activates pending changes.

Two things worth knowing about the PCC. First, Workday is precise that the dashboard data is usage data and not customer content, and that the console itself is not part of the Workday Service. Read that as: it is a courtesy meter, not a contractual one. Second, and this is the one that matters, any customer with a Workday tenant can now turn on the PCC to view document storage. Flex Credits visibility, both complimentary and purchased, is UMSA only.

So you can turn the PCC on and still see nothing useful about your credit burn if you have not signed the UMSA. The gate reaches all the way down to whether you can even watch your own consumption.

Six domains for the ASOR. One for the PCC. Seven security configurations standing between you and a working, observable agent stack. None automatic. That is the gap between what is documented and what is real.

3. The production change nobody scheduled

This one is buried in a feature release note, and it belongs on the first page of every change advisory board deck in the ecosystem.

When you sign the Flex Credits and Platform Entitlement Policy, Workday automatically enables Sana Core for Workday for your organization. As part of that, Workday implements a new native authentication login and password experience in your Production environment. New login UI. New authentication URLs.

And this update happens outside the previously published native authentication rollout schedule.

The act of signing a consumption policy triggers an unscheduled change to how your users log in to Production. If your security team, your help desk, and your comms have not been briefed, you have a Monday morning incident waiting to happen, and the root cause will be a signature on a billing document that nobody connected to authentication.

This is the single highest-value thing in this entire piece. The legal unlock and the production auth change are the same event. Plan them as one.

4. The rate card, decoded

Now the part everyone wants and almost nobody has modeled.

First, in summary, what you get for free. Complimentary Flex Credits, renewed annually, scaled to company size. For HCM or FIN customers under the new policy:

That 200,000 looks generous for those big employers. Hold that thought.

Here is the rate card translated into plain practitioner terms. Every one of these is a real consumption meter from the v262 card*.

*The chart is not the full rate card, just a summary.

The meters are denominated in completely different units. A self-service action and a talent rediscovery run are both “one event,” and one costs 1 credit while the other costs 750. That is the whole game.

A credit is not a credit

Take that same 200,000 complimentary allotment from the top tier. Watch what it buys depending on which agent you switch on.

Same pool. Two-hundred-thousand-to-one spread between the cheapest and most expensive thing you can point it at. The complimentary allotment is built to feel abundant, and for low-cost exploratory skills it is. For high-cost skills it is a rounding error.

The sleeper that should scare you

Payroll Data Monitoring. Five credits per 10 worker records per run. That reads as cheap. Do the unit math and it is 0.5 credits per worker, per run.

Run it across a 100,000-worker population, one run is 50,000 credits. Schedule it weekly, the way you would actually run missing-data monitoring, and you are at 2.6 million credits a year against a 200,000 allotment.

You exhaust the entire annual complimentary pool in four runs. The first month.

Even a monthly cadence is 600,000 credits a year. Three times your allotment, from a single agent configuration that a payroll admin could switch on without anyone modeling the cost. This is not a contract negotiation team carefully redlining 400 documents. This is one scheduled job quietly running against your full headcount.

Model the per-worker, per-run skills before you turn them on. Those are the ones that scale with your population, and population is the one number that does not flatter you.

The platform draw nobody connects to agents

The same credit pool covers platform consumption overages. Core Platform API requests bill at 60 credits per 10,000 calls once you exceed your baseline entitlement, which is 6.5 million calls for the top tier.

Go one million calls over baseline and that is 6,000 credits. Ten million over and you are at 60,000 credits, nearly a third of your complimentary allotment, gone to integration traffic that has nothing to do with agents. Modified Workday integrations and non-Workday integrations count toward the limit. Prebuilt unmodified connectors, Built on Workday apps, and traffic between Workday and the acquired platforms like HiredScore and Adaptive do not.

Your agent strategy and your integration architecture draw from the same well. Budget them together or get surprised together. One timing note: this is the one drain the grace period covers. API overages are waived for policy signers through January 31, 2027, which is exactly why you should use the window to find your real API baseline before the meter starts.

The prerequisite tax

A credit balance is necessary, not sufficient. Several agents require an underlying SKU before you can spend a credit on them. The Talent agents are the clearest example. Recruiting, Talent Mobility, and the Fetch skills all require HiredScore AI for Talent. So the most expensive skills on the card, 750 credits a pop, also require a paid SKU underneath them. You pay to unlock the right to burn.

Check the prerequisites doc before you build a business case on any agent. The credit cost is only half the bill.

5. The grace period and your move before January 31

Here is the urgency, and it is the part everyone is about to get wrong.

There is a grace period. It is real, it is published, and it is narrower than the hype will tell you. Per Workday Community (the Simplified Platform Access post, May 2026), customers who have signed or will sign the Flex Credits and Platform Entitlement Policy pay no overage charges on Application APIs from May 30, 2026 through January 31, 2027.

Read that scope twice. Application APIs. Not agents. Not your Flex Credit burn.

So here is what the grace period actually does. If you blow past your Platform Entitlement for API calls while you tune integrations and agent traffic, those API overages are waived through January 31. That is genuinely useful, because Workday just simplified the entitlement model down to APIs only. Document Storage and Integration Events came out of the Platform Entitlement Policy entirely. Document Storage is now a flat 10 terabytes for every customer. APIs are the only metered platform component left, and the grace period gives you room to find your real API baseline before the charges start.

Here is what the grace period does not do. It does not waive your agent consumption. Run the payroll monitoring job from a few sections back, burn 2.6 million Flex Credits against a 200,000 allotment, and that overage is billable today. The grace period does not touch it. Your complimentary credits let you explore agents in production up to your allotment. Past that, the agent meter bills from day one, grace period or not.

The takeaway most people will miss: the free room is on the platform side, where it is easy. The real money is on the agent side, where there is no grace at all.

And the window only helps if you can see your consumption. Which loops back to setup. No UMSA, no credit visibility in the PCC. No PCC, no meter at all. No ASOR, no agents to meter.

Your move, in order:

1. Check your agreement. MSA or UMSA. Thirty seconds on Community. If you are MSA, start the UMSA opt-in conversation with your account team now, because everything else waits on it.

2. Brief the auth change before you sign. Signing flips Sana Core on in Production. Loop in security, identity, and the help desk first. Make the login change a planned event, not a discovered one.

3. Stand up the ASOR. Six domains. Every tenant. No migration. Start in a sandbox so the runbook is solid before Production.

4. Turn on the PCC. One domain. Confirm you can actually see Flex Credit consumption, which means confirm the UMSA went through.

5. Know your exposure. Use the window to find two numbers, not one. Your API baseline, which the grace period protects until January 31. And your agent burn, which it never protected. Model the per-worker and per-requisition skills hardest, because those are the ones that scale with your size. Walk into February with both numbers, not a hope.

The grace period is a gift with a deadline and a narrow scope. The scope is the part people will miss.

6. What is permanent and what is noise

You know my thesis. Vendor product names have about a nine-month shelf life. The architecture underneath is the thing worth learning.

So separate the two here. The specific rates on the v262 card will change. They are marked confidential and the design-phase skills carry an asterisk that says functionality and pricing can change or vanish at any time. Do not memorize that 750. Do not build a deck around the number 200,000.

What is permanent is the shape. Workday moved from buy-a-SKU-get-unlimited-use to metered consumption, where you pay for the work the agent performs and the meters are denominated in wildly different units. That model is not going back in the box. The names will churn. Sana, Illuminate, whatever comes next. The consumption architecture is the new floor.

Learn the floor. Use the window to learn it with real data instead of marketing math, and learn both numbers: the API baseline the grace period protects, and the agent burn. The clock started May 30. It stops January 31.

And before you click the X in the upper right of your screen, open a new browser tab and go look at your agreement field.

- Mike.

Special thanks to Jess for reading the meter out loud before the rest of us thought to look. Notes, modeling, and any errors here are mine.

tldr;

My 12-year-old figured out anchor-high-discount-down pricing in ten seconds, and "Dad time" as a real cost line in ninety.

Workday Flex Credits work the same way: the rate card is knowable, your actual bill isn't — until you're in production with real users.

Test environments don't consume credits, so your test data is lying to you about the cost.

Practitioner move: run a production pilot before you sign, model the agents that will get adopted (not the demo path), negotiate the flex mechanism not the flex number.

The kid brain and the practitioner brain run the same algorithm. The buyer brain has been trained out of it.

So I posted last week about Justin's first big 3D print order: sixteen tables for the tennis club. The P.S. mentioned that he'd correctly identified labor as a sales lever and was now pricing in Dad time on filament loading and printer support.

A bunch of you wrote in about that P.S. Which is hilarious, because I spent 1,800 words on practitioner confidence and you all wanted to talk about the 12-year-old pricing his father into the deal.

Fine. Let's talk about it. Because the next forty-eight hours got even better.

The Pricing Conversation

After the order came in, Justin sat down to figure out his price for future orders (this one was sold at cost, which is also a correct sales lever to test your potential market. Told you…this kid is scary sometimes).

Materials per table: $30.55.

His first instinct: charge double. Sixty bucks. "Two times cost" is the rule of thumb every tween-with-a-side-hustle defaults to, and honestly, it's not wrong as a starting point.

I pushed back. Sixty doesn't account for failed prints, electricity, printer wear, or his time. The real cost per unit is closer to forty or forty-five once you stack everything in. At sixty, his margin survives one failed print (we’ve had many) and that's about it. I told him to look at seventy-five.

He thought about it for a bit.

"Yeah, I'll do seventy-five. If somebody really wants one and can't afford it, I can come down. That feels better than starting at sixty and having to go up."

I just stared at him.

The Move

That's pricing strategy. That's anchor-high-discount-as-a-lever. That's the move every SaaS rep uses on you in Q4 and you're somehow still surprised when they "find some flexibility" on the renewal call. He landed on it in the time it takes to microwave a mini bag of Pop Secret.

The reason it works isn't magic. Humans hate price increases more than they like price decreases. Going from $60 to $75 feels like betrayal. Going from $75 to $65 feels like a gift. Same fifteen-dollar gap, completely different emotional response.

Justin hadn't read a behavioral economics paper. He hadn't sat through a sales training video. He’s in sixth-grade Social Studies. He just thought about how it would feel to be the person on the other side of the conversation.

The Cost Conversation

The follow-up was sharper.

I started walking him through why $30.55 wasn't his real cost. Filament, sure. But also failed prints. Electricity running four printers. Eventual nozzle replacements on the equipment. Build plate wear. The time he spends slicing, post-processing, gluing, and cleaning up.

His response was, instantly, "What about Dad time? You're loading filament for me."

Reader, I had not raised the topic of Dad time. He just understood that anything keeping the printer running was a cost.

He'd grasped, in about ninety seconds, what most enterprise buyers don't: the unit you're modeling is almost never the unit you're paying for.

Which brings us to Flex Credits.

What Justin Just Figured Out About Workday Agents

If you haven't had to think about Workday Flex Credits yet, you're about to. They're the consumption-based pricing model for Workday Agents (the Sana-based AI tools rolling out across the platform). You don't pay a flat fee. You pay for what you use. Test environments don't consume them. Traditional ML tools don't consume them. They're specifically for Agents, in production, with real users hitting them.

Which means here's what you actually know before you sign:

• The rate card.

• Your headcount.

• Your guess at usage.

Here's what you don't know:

• Which agents your population will actually adopt.

• How often they'll hit them.

• What a "popular" agent looks like at your scale.

What the bill looks like in month four when somebody builds an agent that gets embedded in three high-traffic business processes.

You can't know any of that until you're in production. The Flex Credits FAQ on Community will give you the structure. It will not give you your bill. The bill is a function of usage, and the usage is a function of behavior, and the behavior is unknowable until the thing is live.

This is the same problem Justin hit with the tables.

He could model materials. He could not model the failure rate until he ran the batch. He could not model his own labor until he was elbow-deep in a four-printer relay. The sticker number ($30.55) was the cleanest, most knowable variable. It was also the smallest piece of the actual cost.

Workday Flex Credits will work the same way. The rate card is the cleanest, most knowable variable. It is also the smallest piece of the conversation you should be having.

The Practitioner Move

If you're building toward an Agents rollout (and if you're a Workday customer, you are, whether you've gotten the signal yet or not) the practitioner version of Justin's instinct is this:

1. Run a production pilot before you commit. Test environments don't consume credits, which means your test environment is also lying to you about your costs. Buy enough credits for one real workflow with real users for one real cycle. That's your ground truth. Everything else is the rate card multiplied by a guess.

2. Pick the agents most likely to get adopted, not the ones easiest to demo. Demo agents are cheap because nobody uses them. The agents that get adopted hit your credit balance hardest. If you're modeling cost based on the demo path, your bill will be wrong in the same direction every time.

3. Negotiate the mechanism, not the number. Anchor on what happens when usage exceeds projections, not on what the projections are. The projections are wrong. They're always wrong. Justin's move (quote high, leave room to come down) is the same move you want from your CSM. Ask for the flex band, not the flex number.

The Bigger Thesis

Here's the part that really bothers the s&!% out of me.

A 12-year-old figured out, in ninety seconds, what enterprise buyers routinely miss in seven-figure Workday conversations. He's not smarter than the buyers. He's just not trained to ignore what's right in front of him.

He doesn't have a procurement playbook telling him "license cost is the cost." He doesn't have a vendor deck telling him orchestration is the future. He doesn't have a webinar pre-loading the answer. All he did was look at the whole picture and asked what it actually takes to get the table out the door.

That's the practitioner brain. First principles. No vendor framing. No domesticated buyer reflexes.

The kid brain and the practitioner brain run the same algorithm. The buyer brain runs a different one (a trained one) and that's how you end up signing a Flex Credits commitment at a number you'll outgrow in month four.

Charge $75. Discount when it makes sense. Run the pilot before you sign.

A 12-year-old figured it out. Your renewal is coming.

— Mike

P.S. Justin has now extended the breaded chicken billing model. Filament loading remains the base service. Post-processing has been quoted at "lemonade rates." I am told these are non-negotiable. The Department of First Things First's first pricing analyst is, apparently, also its first finance director.

tldr; The Salesforce envy is real, but the gap isn't celebration (or the lack thereof). It's leverage. The Workday practitioner community already exists (it's just self-organized by necessity). The vendor is sending mixed signals. The deeper sting is structural: in our ecosystem, the architect and director roles go to consultants, not practitioners. We don't fix that by waiting for a jacket. We fix it by self-organizing harder and using the leverage we already have. Flex Credits proved it works.

I'll be honest with you. I get the FOMO.

When I scroll past my Salesforce friends collecting Trailblazer badges, stacking certifications like Pokémon cards, wearing the rainbow jackets to Dreamforce, there's a part of me that thinks: "why don't we have any of that?"

We get Workday Pro. Which is a transactional cert program. Useful, marketable, but not exactly an identity. There's no jacket. No campfire. No "I'm a Workday-er" t-shirt I'm putting on my LinkedIn header.

The Nick Moores promotion was a signal that maybe (maybe) that's starting to shift (for those of you not in the know, Nick has been a developer advocate for Workday since 2021, and an integrations consultant before that. He's been quietly been building a presence for years. He knows his stuff).

Then, in the 2026 layoff round, they cut their community manager.

So the signals are confused. But honestly? That's not the part that matters.

The community already exists. The vendor just hasn't shown up (fully).

Look around. The RUGs. The Customer Sharing Movement. The Unsafe Harbor podcast. Workday Gold. HRIS Careers. And that's the tip of the iceberg. There's so much more good, solid content that practitioners are putting out there. The DMs that fly between practitioners every time a release note drops. The Slack channels nobody talks about publicly (I have a WhatsApp thread from Rising 2025 that's still alive - Go Social Squad ‘25!). The conference hallway conversations that are more useful than the keynotes (I once had a brilliant conversation about building a Slack integration with Workday two full years before Workday did in a Las Vegas elevator. Again, Eric, thank you. You’ve truly been a friend ever since).

The ecosystem is bottom-up by necessity. The vendor didn't build it. We did.

It's actually pretty good. Quieter than Trailblazers. Less manufactured. Fewer chants. The energy is operators talking to operators, not vendor-led PR cosplay. There's a kind of dignity to that.

But the gap is real. And it isn't really about jackets.

The real sting: who gets the title?

In our ecosystem, the "Solution Architect" and "Director HR Technology" roles disproportionately go to consultants who rotated through six tenants and built a personal brand on LinkedIn. Practitioners (the people who've actually lived in one tenant for eight years, who know where the bodies are buried, who've made the system survive three CHRO changes and a reorg) get a Workday Pro cert and a LinkedIn post when they get their certification.

That's a leverage gap.

Salesforce admins have an ecosystem that translates community participation into career mobility. Recruiter pings. Conference badges that mean something on a resume. A clear path from "I work in Salesforce" to "I architect Salesforce." That path exists for Workday consultants. It barely exists for Workday practitioners.

So when the Trailblazer FOMO hits, that's actually what we're feeling. Not the jackets. The economic ecosystem the jackets are downstream of.

The good news: we have more leverage than we think.

The Flex Credits walk-back proved it. Practitioners pushed back. Publicly and persistently. The vendor moved. Let's call it a market correction triggered by practitioner pressure.

The lesson is: when the practitioner community speaks with a coordinated voice, the vendor listens because they have to.

So here's the actual ask:

Stop waiting for the jacket. The jacket isn't coming, and frankly, after looking at the Trailblazer ecosystem up close, I'm not sure I want one.

What I want is the vendor to back the community that's already here. Sponsor the practitioner-led podcasts. Send Nick to RUG meetings. Promote practitioners into architect and director roles, not just consultants with implementation badges. Stop treating community as a marketing function and start treating it like a product input. And for the love of all things proper, BRING BACK BRAINSTORMS.

And until they do? Keep self-organizing. Show up to your RUG. Subscribe to the practitioner newsletters. Quote each other. Recommend each other for roles. Build the muscle the vendor hasn't built for us.

We don't need the jacket.

We need the leverage. And we already have it.

P.S. I asked Justin if he wished his volleyball team had cooler uniforms. He said, "the team that wins the championship gets a banner at the club. The jerseys don't matter that much." Then he asked why grown-ups care about jackets. I didn't have a great answer.

— Mike

Director HR Tech | No Jacket Required

The Department of First Things First. For the people who do the work.

tl;dr

My son Justin just completed an order for 16 custom 3D-printed side tables for our tennis club. He almost passed on it before he started, because he didn't think he could do it.

Once he got out of his own head, he did exceptional work: researched a Creative Commons base design, modified it for tennis (added ball holders so balls don't roll away), and proposed adding the club's logo. The owner ordered 16 on the spot.

The framework that closed the deal is the same one that closes any deal, in any industry. Listen to the actual problem. Show up with a working solution. Add the detail nobody asked for. The hardest part is usually getting yourself to start.

Our tennis club just installed Zenniz line-calling on every court. If you don't follow tennis tech, it's basically Hawk-Eye for amateur clubs. Cameras, AI, automatic line calls. It's slick, it works, and it's almost entirely killed the line-call argument that has been ruining doubles matches since the invention of doubles matches.

It also killed our courtside tables.

The little tables that used to bolt onto the net post had to come off to make room for the main system mount. So now there's nowhere to put a water bottle, a phone, or a tube of balls between sets. Players were grumbling. The owner was hearing about it constantly. Nobody had a good answer (I know, first world problem. I fully acknowledge my privilege).

Jenn and I were talking about it on the way home from a clinic, and she said the obvious (AKA smart) thing.

"Why doesn't Justin 3D print one?"

She floated the idea to the owner. He was interested. One of his juniors was going to help out the club. He didn't need a deck. He didn't need a roadmap. He needed a table that could survive a tennis club.

Then we told Justin.

The Part Where He Almost Bailed

Justin's first reaction was not enthusiasm.

He actually screamed "I can't do that,” and promptly hung up the phone on his Mom.

Now, let’s put his reaction into perspective. He runs a four printer farm in our basement. He's been modeling in Fusion 360, Tinkercad, Canva, and Bambu Studio since he was ten. He has produced more functional, finished objects in the last year than most adults will in their lifetime. None of that mattered in the moment. The minute it became real, with a real customer expecting a real product, his brain immediately went to all the reasons he wasn't the right person for it.

It was too big. He'd never done a real order before. What if it didn't fit together? What if the panels warped? What if the owner hated it? What if, what if, what if.

This is, by the way, a very specific feeling that I have watched grown adults experience (me too!) in conference rooms hundreds of times. Same posture. Same furrowed brow. Same internal monologue. "I'm not the right person for this." Mostly delivered by people who, by every measurable standard, are exactly the right person for this.

This time, it took one conversation once we got home. Some "you have literally already built harder things than this." Some "the worst case is the owner hates it and we eat the filament cost." Some appeals to the obvious upside (real money, his name on a product in a real business).

And then, somewhere between the third and fourth point, the imposter syndrome shut off and the maven flipped on.

What He Actually Did

From that point forward, he was sprinting.

He found a Creative Commons table design he liked as a base (because the smart move is never to start from scratch when somebody has already solved 60% of the problem for you). He modified the dimensions to work as a courtside companion. He sized it to drag around for tournaments without throwing your back out, but strong enough to be knocked around and not break.

Then he made two decisions that I want to flag, because they are the entire ballgame.

First, he added circular ball holders to one of the panels.

Three little wells, sized to hold a tennis ball, so they don't roll off the table the second you set them down. Anyone who has played tennis for more than fifteen minutes knows this is the single most relatable design problem in the sport. Tennis balls are round. Round things roll. Tables are flat. Flat plus round equals annoyance. Justin solved for that, on his own, because he plays the sport and he has personally experienced the exact moment of frustration the table needs to fix.

Nobody asked for ball holders. He just understood the use case.

Second, he proposed the club's logo on the top.

Not a sticker. Painted on with a machine. Multi-color, right on the panel. The owner had not asked for any branding. Justin offered it because, on some intuitive level, he knew that a generic side table is something you order from Amazon. A side table with the club's logo on it is something the club owns.

The logo is the difference between a piece of furniture and a piece of identity.

He built the prototype. Hexagonal top, triangular panels in the club's colors, structural base in black PETG (stronger plastic to survive being knocked around), ball holders, logo on top. Walked it into the pro shop. The owner picked it up, set a Gatorade on it, looked at the logo, and ordered sixteen.

A sample is in the photo (above). The last panel is currently out at a vendor getting the logo printed. The basement smells like warm plastic. There is a four-printer relay running in shifts. Toby has assigned himself to quality assurance and currently sleeps on the couch next to the “farm”. We are operating at full tilt, and Justin has, in the last 72 hours, gone from "I can't do this" to texting me production schedules.

It is a hell of a thing to watch.

Share

The Three Things That Closed the Deal

This whole story comes down to three moves. They are stupidly simple, and they apply whether you are selling tables, software, or services.

1. Listen to the actual problem

Not the problem in a category report. The problem the customer is actually complaining about, in actual words, in the actual room.

Jenn heard it first. The Zenniz install created a real, specific, daily friction. Players had nowhere to put their stuff. The fix didn't need to be sophisticated. It needed to be a table.

How many vendor demos have you sat through where the AE has clearly never read your annual report? Where the discovery call was a checklist of features they wanted to demo, not questions about what you actually need? Where they show you the agentic AI demo when you said the meeting was about merit planning? The reason most enterprise software pitches fail is not that the product is bad. It is that nobody bothered to find out what the customer was actually trying to do.

2. Show up with a working solution, not a roadmap

The owner was interested in the concept when Jenn mentioned it. But the order didn't happen until Justin walked in with a thing the owner could pick up.

That is a profoundly different sale. The buying decision was not "do I trust the roadmap." The buying decision was "do I want sixteen of these."

The HR Tech version of this is the team that walks into a steering committee and says "I built a working prototype of the thing you asked for, here is the screen, here is what happens when I click this button" versus the team that walks in with a Visio diagram, a list of dependencies, and a Q3 milestone called "TBD."

Which team gets funded? You already know.

3. Add the detail nobody asked for

Logo. Ball holders. Pick your favorite.

The logo is the obvious customization (branding makes it feel owned). The ball holders are the non-obvious one, and they are the more important one. The ball holders only exist because the person designing the table actually plays the sport and has personally lived the moment of frustration the design needs to solve.

That is practitioner empathy (Thanks Jerome!). It does not come from a requirements document. It does not come from a workshop. It comes from someone who has been in the seat the customer sits in.

In Workday terms, the logo is the embedded help text and the org-specific landing page. The ball holders are the business process change you made because you knew the manager was going to bounce out at step seven if you didn't change that approval to a notification. Both matter. The second one is what separates implementations that get adopted from implementations that get tolerated.

The Fourth Thing, Which Almost Killed the Deal

The framework above describes what Justin did. It does not describe the part that almost stopped him from doing any of it.

He almost said no.

Before the design. Before the research. Before the prototype. Before the order. There was a moment where the entire deal hinged on a 12-year-old's willingness to believe he was the right person for the work.

If a 12-year-old with a 4-printer farm can see the path through the fear, why are we, with 20 years of logic-based experience, still stalling? You have sat on an idea you knew was right because "that's not really my call to make." You have deferred to a consultant on something you knew more about than they did. You have watched a vendor pitch a half-baked solution to your CHRO and not pushed back, because who are you to push back? You have written a draft of a recommendation that you knew was correct and never sent it, because, “what if you were wrong?”

The work was never the hard part. The hard part is always the moment before the work, where you decide whether you are the kind of person who does this thing or not.

The HR Tech industry has a chronic, unaddressed practitioner-confidence problem.

The people closest to the system, the ones who actually understand the data model, the security framework, the manager experience, are constantly deferring to people who are paid more, sit higher, or talk faster. Meanwhile, the analysts publish, the consultants bill, the vendors demo, and the practitioners stay quiet. We treat "I'm just the admin" as humility when it is actually the single biggest blocker to better systems.

Justin almost talked himself out of his first big sale. You have probably talked yourself out of your last good idea. That's the bigger story.

The Bottom Line

The framework is real and it works. Listen, prototype, add the detail nobody asked for. You will close more deals, internally and externally, than the people who don't.

But the framework only matters if you actually start. And the part that nobody warns you about is that starting is the hardest part of every meaningful piece of work you will ever do.

My 12-year-old got over it in one conversation. Some practitioners I know have been stalling on the same idea for three years.

Be more like Justin. Once he flipped, he did not look back.

— Mike

Director HR Tech | Production Support, Domingo-True Lab

P.S. Justin asked me last night what his “profit margin” should be. I told him to factor in materials, electricity, printer wear, and his time. He thought about it for a few seconds and said, "What about Dad time? You're loading filament for me." I told him he could pay me by playing doubles this weekend with me. He said "deal" and went back to slicing the next batch. The kid who, four days ago, didn't think he could do this is now correctly identifying labor as a sales lever and pricing in his father. The Department of First Things First just hired its first pricing analyst. He bills hourly and accepts payment in breaded chicken cutlets.

P.P.S. This was a huge Proud Dad moment for me. So of course, I made a video of production. If you're interested, here's the video:

The Department of First Things First. For the people who do the work.

tldr;

Sana AI Summit on May 21 was a coronation, not a product launch. Aneel publicly handed Workday's AI future to Joel Hellermark, the 28-year-old self-taught Swedish prodigy who is now Chief AI Officer.

Aneel re-staged Dave Duffield's 2016 "don't screw it up" line. In front of customers. Twice in a decade. Workday's own copy called it a "refounding."

Sana for ITSM is the ServiceNow shot. Sana for Travel and the agent demos showed real agentic workflows for hiring, access requests, and expense filing.

The agent layer is a new governance attack surface. Agents are not just doing the work. They are learning your approval patterns and proposing to rewrite your policies.

The new Flex Credits rate card quietly dropped or deferred integration events, documents, and storage from the meter. A customer-friendly retreat. The new pricing risk now lives entirely in agent consumption.

What you do Monday: track Joel not the brand, treat agents as a governance object, get the agent consumption assumptions from your AE before your next renewal.

I told Justin I was going to a tech summit in New York. Sana. Workday's new AI brand. Big day, big speakers, fancy library.

His response was pure gold.

"So you were hanging with billionaires."

Yeah, son. I guess I was.

That's the right framing for what happened at the New York Public Library on May 21. Not a product launch. Not a roadmap update. A coronation.

Act I: The Hand-off

The day opened with the 1977 Powers of Ten film. Chicago picnic blanket to the observable universe and back. Designed to make you feel small.

That was on purpose. Joel Hellermark, founder and CEO of Sana, wanted you small. He wanted you small so the next part felt big.

Three acts. The first one is the only one practitioners need to study.

Aneel Bhusri walked on stage in a Björn Borg jacket. The Swedish apparel brand. Worn by the co-founder of an American enterprise software giant being interviewed by a 28-year-old Swedish AI prodigy he just acquired. The jacket is the slide.

He sounded somehow older than I remembered. Gravelly. He acknowledged it himself. Said he was a different guy than he was at 30. Said he now felt like an orchestration agent for Workday the company.

Joel did the interviewing.

The new guy hosted the founder. That is the entire blocking diagram of who is running the next chapter.

Late in the conversation, Aneel delivered the line that justifies the whole bet. Reasoning engines, he said, are weak at deterministic plumbing. The power comes from bringing the two technologies together.

Translation for the practitioner audience: Sana without Workday is a smart toy. Workday without Sana is yesterday's deck. The merger is the thesis. If you ever wondered why Workday bought a Stockholm learning company, that's why.

Who is Joel Hellermark, and why should you care

Short version: he is the closest thing enterprise software has to a generational prodigy story.

Taught himself C at 13. Built a chess program for fun. Founded his first company at 16. Skipped university entirely. Learned to code from Stanford's free online courses. Founded Sana at 19 out of a converted hair salon in Stockholm. Named Sweden's "Super-talent of the year." The Guardian's 10 under 35 changing the world. Aneel calls him "one of the most brilliant minds in AI."

He is now in charge of Workday's AI future.

Internal title: Chief AI Officer. Internal mission, in his own words, is to be the Hermes of Enterprise AI. No more bad UX. Software you actually want to use.

"Don't screw it up"

Late in the Act I interview, Aneel got quiet.

He reminded Joel that at Workday Rising 2016, Dave Duffield said the same four words to him. In front of customers.

Don't. Screw. It. Up.

Now Aneel was saying it to Joel. In front of customers again.

This is the moment to stop scrolling.

The line was a public ritual in 2016. Aneel just re-staged it. That is not a coincidence. Founders do not perform their predecessor's parting line by accident. They do it when they want the room to witness a transfer of authority. The company Aneel built is being handed to a brilliant kid who was not born when Workday's first customers signed.

Aneel called this a refounding moment. The pre-event LinkedIn copy used the same word. A complete re-founding of how work gets done.

Their word. Not mine.

When the vendor uses the word refounding, what they are telling their customers is: the company you bought from is being rebuilt while you are still inside.

The three things you actually need to know

Strip away the philosophy, the GDP forecasts, and the closing meditation on what stays human. Here is what changed on May 21 for the people turning the wrenches.

One. Joel is the heir. Not co-leader. Not strategist. The face of the next decade. If you are building governance, integrations, or roadmap plans against Workday, the person whose taste matters most is now the one who learned coding from Stanford YouTube videos. The branding will keep moving. His taste is the constant. Track him, not the press releases.

Two. The agents are real and they are coming for your queue. Sana for ITSM and Sana for Travel were both announced. Both are agent-led, not feature-led. Demo flows showed the agent brokering a Salesforce admin access request, drafting an offer letter for a $151,000 SGD Senior AE in Singapore, and booking a Stockholm trip while filing the expense in Workday in one motion. Three pillars Joel kept returning to: multiplayer interface, living memory, proactive agents.

Three. Workday is going after ServiceNow. Sana for ITSM means Workday is no longer staying in the HCM and Financials lane. Fulfill requests. Automate ticketing. Manage assets. Build workflows. That is ServiceNow's deck rendered in Workday's typeface. Your platform is now a horizontal competitor to a different ecosystem entirely. I think Workday is going try and move up and down the big business tech stack as much as possible.

The slide that should keep you up at night

The Sana for ITSM demo had a Salesforce access flow. Kai Tran, a Sales Lead in APAC, needed admin rights to set up new-hire territories. The agent scoped it down, surfaced the precedent (same scope Marcus has in EMEA, approved 12 months ago without escalation), and asked the approver to sign.

That is a credible agent demo.

The next slide is the one I want every HRIS lead and access governance owner to look at.

The agent comes back and says: You always approve these. 8 of the last 8 CRM admin requests from Sales leads were approved without changes. Want me to take regional CRM admin off your plate going forward?

Then it offers a draft policy update. Auto-approve when a Sales lead requests their own region's CRM. Escalate when scope crosses regions. Reversible toggle.

The agent is not just doing the work. It is learning your approval patterns and proposing to rewrite your governance.

This is a new attack surface for HR tech and access governance. Not malicious. Just new.

If your access policy can be amended by an agent recommending the change based on observed behavior, then your approval queue is now training data for your future policy. Practitioners need a vocabulary for this before the audit team does.

Share

What they did not say

Day-long thought leadership session. No Flex Credits discussion from the stage.

Which is interesting, because Workday actually had good news to share.

Go read the new Flex Credits Platform Policy rate card. It will be published next week. Compare it to the model they rolled out at Rising 2025. Integration event overages are deferred to 2027. Document events are gone. Storage is gone. The three buckets the analyst class spent six months calling a budget black hole have been quietly pulled off the meter.

That is a customer-friendly retreat. Workday heard the pushback and walked it back.

So why didn't they say it from stage?

Because the new meter is a single-bet meter. Everything you already do at scale, the integrations you run nightly, the documents you store, the events you fire, is now off the table. What's left on the meter is agent consumption.

The thing your tenant cannot yet size, because your tenant has not deployed it.

Workday swapped a known-quantity risk you could model in a spreadsheet for an unknown-quantity risk you cannot model until the agents are live. That is a real tradeoff and arguably a better one for most customers. But the FinOps muscle required to land it is different. You are no longer sizing integration volume against a credit pool. You are sizing how often a Recruiting Agent will work a pipeline and how often an IT Support Agent will pre-broker an access request and how often a Travel Agent will book and file an expense in one motion.

That conversation does not happen from a stage. It happens in a sales motion with your AE.

Other things that did not get said from the stage:

- No migration path for customers already invested against the Illuminate naming

- No customer reference architecture for Sana for ITSM

- No deployment story for how Sana actually grounds against your Workday tenant

- No GA timing beyond the headline launches

The analyst class will spend the next two weeks writing about what Workday announced. Most of them will not go find the rate card.

Acts II and III, briefly

Act II zoomed out. Anton Osika from Lovable on vibe coding democratization. Tyler Cowen on AI as radical evolution with a 2.5% annual GDP growth rate, exceeding baseline forecasts. Then the panel's dark moment. Jasmine Sun described standing in front of a humanoid robot in China and feeling viscerally that this thing could kill her. Benjamin Labatut, the only pessimist on the stage all day, asked why we are not paying more attention to that feeling. Nobody had a good answer.

Act III went philosophical. Geoffrey Hinton in person, not virtual (I am officially walking back my virtual stage comment). AI may outstrip mathematicians inside a decade. We are creating beings and optimizing for intelligence when we should be optimizing for care. Sara Imari Walker on assembly theory and what makes something alive. Anu Athluru closing with the question. What stays human when machines can do the work? Her answer was seven things. Taste. Grit. Care. Craft. Story. Purpose. Meaning.

Soul was the descriptor on the final slide.

It was, in case it wasn't already obvious, a very European summit. Swedish CEO. Swedish prodigy. Swedish brand on the founder's back. American enterprise software's center of gravity just shifted across the Atlantic by about 4,000 miles.

It rained all day outside the library. Inside, it was a TED talk sponsored by Workday. Light on product, heavy on framing. That is a tell too. When you cannot show the product working outside a carefully crafted demo yet, you show the philosophy.

What you do Monday

You do not change anything Monday. But you start tracking three things.

One, watch what Joel says, not what Workday says. Vendor branding has a shelf life of about nine months. Joel's taste does not. He is the new center of gravity.

Two, treat the agent layer as a governance object, not a feature. Every agent that can rewrite a policy needs an owner, a kill switch, and an audit trail before it ships into your tenant. If your governance forum does not have agent governance on the agenda by Q3, you are late.

Three, before your next renewal, ask your AE for the agent consumption assumptions baked into your Flex Credits sizing. Not the rate card. The assumptions. How many agent operations per user per month. How they modeled it. What happens at 2x. If they pause, you have learned something useful.

Aneel said this is a refounding moment. He is absolutely right. The company you have been buying from for ten or fifteen years is becoming a different company.

The kid who didn't go to college is in charge of getting it there. The founder who built the cloud HCM category just told him don't screw it up.

The rest of us go back home to a tenant we still have to run on Monday. Same configurations. Same approvals. Same governance committee meetings.

For now.

So what stays human when machines can do the work? Anu's answer was taste, grit, care, craft, story, purpose, and meaning. Mine is shorter.

The judgment about whether to let them.

— Mike

P.S. — I told Justin about the day on the train ride home. About Aneel and Joel and the Powers of Ten film and Hinton not being virtual (he did think that was cool).

He didn't ask about any of that.

He asked if the food was good.

That's about right.

tl;dr;

I got an invite to the Sana Summit on May 21. I'm going. The room will be Geoffrey Hinton, Tyler Cowen, Ethan Mollick, Aneel Bhusri, and somewhere in the back, me. Most of the recaps Friday will read like Davos liner notes. I'm publishing this before the event so I'm on the record about what I refuse to be impressed by (and what I'm actually watching for).

I had to read the email twice.

The Sana Summit. May 21. The New York Public Library. Geoffrey Hinton on the (virtual) stage. Tyler Cowen. Ethan Mollick. Sara Imari Walker. Aneel Bhusri.

And somewhere in the back of the room, me. A Sr. Manager from the basement of an HRIS team at a Fortune 10, holding what I assume will be a name badge in a font I cannot read.

I don't fully know why I got invited and I'm not going to pretend otherwise. Maybe the newsletter. Maybe I was just lucky. Maybe the RUG. Maybe CVS is a big enough customer that at least one of us was going to end up there. Maybe someone in Workday's orbit reads this and decided the practitioner column deserved a chair. Probably some of all of that.

What I do know is what I'm bringing into the room.

The product story already shipped. In March, Sana for Workday went live worldwide. Sana Self-Service Agent. Sana Enterprise. All priced through Flex Credits. Bersin already wrote his big take. Sapient too. Twelve hundred customers are already registered on the Agent System of Record.

Thursday is not a roadmap event. It's a consecration. Workday is buying philosophical legitimacy by putting Aneel Bhusri on a stage with a Nobel laureate. Every vendor does this when they want to graduate from "enterprise software" to "intellectual authority on the future of work." The photos will be tasteful. The recaps will be reverent. The analyst class will spend a week quoting it.

Most of the room will be fine with that. I am going to take notes anyway.

So here it is. The pre-commitment.

Here's what I won't do.

I won't post selfies with Hinton. (Probably won't get close enough to anyway.)

I won't write a recap that opens with "what a privilege to be in the room with such brilliant minds." If you see that sentence in a newsletter Friday, the writer has been captured.

I won't let Tyler Cowen's productivity numbers sit unchallenged when they map poorly onto an actual Workday tenant.

I won't nod through the Aneel Bhusri keynote without asking the questions the keynote will not answer.

Here's what I will do.

I'll watch the gaps. What's not said about Flex Credit math when an agent pilot ramps to production. What's not said about the security model when Sana Enterprise reads and writes across Slack, Salesforce, SharePoint, and Box at the same time — all under the Workday security groups you've been meaning to clean up since 2022. What's not said about who owns the governance review the day an agent recommendation breaks comp equity six months after twenty people approved it by not questioning it.

I'll do my best to bring the basement into the room. Not loudly. Just in the questions I ask and the parts I choose to remember.

And Friday, when the consecration recaps are everywhere, I'll write back. From inside the room. Un-consecrated.

The cool kids' room is in midtown Manhattan Thursday afternoon. Surprisingly, I'll be there.

The actual work is in the basement, where it's always been. I'll be back there Friday.

— Mike

Practitioner Un-consecrated

The Department of First Things First. For the people who do the work.

P.S. Justin asked me what I was doing this week. I told him about the conference. He asked who was going. I told him Nobel winners, Wharton professors, and his Dad. He looked at me for a second and said, "Wait - they invited YOU?"

Same question I had, kid. Same question I had.

• Last week I said Workday is becoming PeopleSoft. This week, what you do about it.

• Some of the most sophisticated Workday customers on the planet have already decoupled their stack this way. Workday underneath as the deterministic engine. An experience layer on top.

• Three layers: the deterministic engine, the integration fabric (where everyone botches it), and the experience layer (where the war is actually being fought).

• Workday knows all of this. Their instinct is to own all three layers themselves, and Flex Credits is the receipt. That’s exactly the wrong move, for them and for you.

Last week’s post did a jawn (that's Philly for a thing).

It also triggered the email I should have expected but didn’t. From a reader:

“Okay Mike. Great thesis. What am I supposed to do with this on Monday morning?”

Fair. Diagnosis without prescription is just complaining in a nicer font.

So here’s the prescription. The short version is decoupling. The long version is three layers, a failure mode at each one, and one specific move you make this quarter. All of it set against the fact that Workday is actively trying to solve this problem, and in my view, trying to solve it the wrong way.

But before any of that, we need to deal with a thought that’s rattling around in your head right now.

Why This Isn’t Theoretical

Decoupling isn’t a hypothetical future-state. The most sophisticated Workday customers on the planet have already done it. Quietly. Years ago. While the rest of the industry was still arguing about whether the native inbox was good.

I’ve confirmed the pattern across three companies.

A former CHRO at a major e-commerce platform walked me through their decoupled stack in a recent (2025) conversation. Workday as the engine. A purpose-built experience layer on top. Employees do not touch the ERP. They touch the front door the company built for them.

A major social platform runs similar architecture. I know because I was interviewed for (and offered) the integration architect role on their Workday implementation by the SI doing the work. I turned it down. I saw the stack. Workday underneath. Experience layer on top. Same pattern.

And Google’s version is public record. gHire is their in-house recruiting front end. It sits on top of their HRIS stack, not inside it. It exists because Google’s recruiters needed a tool built for recruiters, not a tool built for compliance officers who happened to work in recruiting.

Three of the most demanding HR tech environments on the planet. Three different industries. Three independent decisions to decouple. Same pattern.

Here’s what they figured out that the rest of the industry hasn’t: the front door was never going to be the deterministic engine.

They knew it ten years ago. The difference now is the rest of us can actually afford to act on it. You don’t need Google’s engineering bench anymore. LLMs collapsed the cost of building a decent conversational front door from “moonshot project” to “quarterly initiative.” The experience layer used to be a capability you had to build. Now it’s a capability you can buy, configure, or compose.

The playbook is the same. The price of admission just dropped by three zeros.

So let’s go.

Layer 1: The Deterministic Engine

This is Workday. It’s also your financials system, if you have one. It’s the load-bearing stuff. Payroll. Benefits. Time tracking. Compliance reporting.

This layer has one job: be right. Every single time. When the paycheck drops, the math is the math. When the I-9 is filed, the form is the form. There is no room for “creative” in this layer. There isn’t much room for “fast” either.

That’s a feature, not a bug. This is the layer that keeps you out of court.

The failure mode here is treating the engine like it’s the whole system.

This is the mistake most of the HR tech industry has been making for fifteen years. We built massive, compliance-hardened transactional databases and then asked the 19-year-old applying for a retail job to navigate them. We asked the store manager with five active alarms to execute a 14-step requisition process on a cellphone, or an old tablet. We handed the engine to people who should never have to touch it.

(The “Moat is a Tollbooth” thesis applies here. The decades of business rules that make the engine powerful are the exact same decades of business rules that make it unbearable for end users. The engine’s strength and the engine’s UX are the same variable, measured from different sides.)

The move this quarter: narrow the engine’s job description.

Take thirty minutes and list every workflow where your end users touch Workday directly. Time off. Pay stub access. Benefits enrollment. Job requisitions. Performance check-ins. Org chart lookup.

For each one, ask a single question: does this workflow belong here because it requires the deterministic guarantees of the engine, or does it belong upstream in something that isn’t here yet?

You’re not ripping anything out. You’re making a list. That list becomes the input for Layer 3.

Layer 2: The Integration Fabric

This is the connective tissue. APIs, middleware, event streams, identity, webhooks, the whole plumbing situation. It’s the layer that lets the engine talk to everything else without the employee ever knowing a handoff happened.

Nobody wants to talk about this layer. It doesn’t photograph well. You can’t put it on a Rising keynote slide. There are no “Top Ten Integration Fabric Innovations of 2026” listicles. It is the unfinished basement of your architecture. (I know all about unfinished basements. I routinely lose to a 12-year-old at ping-pong in one).

Which is why it’s where most orgs quietly die.

The failure mode here is accumulation. Not design.

You didn’t decide to have 47 point-to-point integrations. You decided to have one. Then someone needed a connector to Greenhouse. Then someone else needed a custom feed to a benefits broker. Then a vendor sunset their old API and you duct-taped a bridge. Then the acquisition happened and you inherited their spaghetti. Now you can’t name the person who owns the feed between your HRIS and your learning platform, because the person who built it retired in 2021.

This is the Invisible Spaghetti problem. I have walked into organizations that genuinely cannot tell you how many integrations are running. They don’t know the number. They don’t know what breaks if half of them die tomorrow. They don’t know which ones are point-to-point and which are orchestrated through middleware, because they don’t have a coherent middleware strategy. They have a middleware accident.

The move this quarter: run the real integration audit.

Not “how many do we have.” The real one. For every integration touching your HRIS, document four things:

What’s it doing. Who owns it. What breaks if it dies. Whether it’s point-to-point or orchestrated.

If you can’t draw the whole integration map on a whiteboard in ten minutes, you have a middleware problem. That’s the project. Not a new AI pilot. Not another vendor demo. The project is getting the fabric readable.

(You cannot orchestrate what you cannot inventory. Write it down.)

Layer 3: The Experience Layer

Here’s where it gets interesting. And where the Workday story gets complicated.

The experience layer is the front door. It’s where employees actually live when they interact with “work.” It might be a conversational AI that answers their policy questions. It might be an embedded flow inside Slack or Teams where they request time off by typing a sentence. It might be a candidate experience platform that handles the entire pre-hire journey before anyone touches the ATS. It might be an agent that watches your calendar and proactively flags PTO conflicts before you create them.

What it is NOT, in a healthy decoupled architecture, is your HRIS with a nicer skin.

The whole point of the experience layer is that it lives above the engine and the fabric. It’s purpose-built for humans, not for compliance. It can be consumer-grade fast because the stuff that has to be deterministic is happening a layer down. The experience layer can rewrite itself every quarter. The engine shouldn’t change in five years.

The failure mode here is putting the experience layer in the wrong place.

Which brings us to Workday.

Workday sees the problem. They are not stupid. They have watched the same twenty years I’ve watched, and they know the ERP-as-front-door model is over. That’s why Illuminate happened. That’s why HiredScore got absorbed. That’s why Sana got bought for a billion dollars. That’s why the keynote narrative at Rising has quietly pivoted from “your system of record” to “your system of action.”

The problem is their answer to the decoupling question is: we will own all three layers.

Sana is being pitched as the experience layer. Illuminate (or whatever it’s called next quarter, because the branding has a shelf life of about nine months) is being pitched as the orchestration fabric. The core tenant remains the deterministic engine. One vendor, three layers, one throat to choke.

And look, I understand the commercial logic. For Workday, decoupling is a revenue problem. If the experience layer becomes a competitive market and the ERP quietly becomes headless infrastructure, Workday becomes a utility. Utilities don’t trade at software multiples. So the strategic imperative is to make sure the experience layer is ALSO Workday. Sana isn’t just a feature. Sana is a defensive maneuver against becoming PeopleSoft.

But the whole reason the three-layer architecture works is that the layers have different incentives, different release cycles, and different evolutionary pressures. The engine wants to be boring and stable. The fabric wants to be observable and governable. The experience layer wants to be fast, opinionated, and disposable. You cannot optimize for all three from inside the same vendor.

When a vendor tries to own all three layers, they always optimize for the engine, because that’s the moat. The experience layer ends up being “our engine with a conversational interface on top,” which is the exact thing decoupling was supposed to fix. It’s a tollbooth with better lighting.

And if you want the receipt that Workday has already privately conceded the engine-only future? Look at Flex Credits.

We covered the mechanics in the Sana Field Guide a few weeks back. Here’s what the pricing model really tells you. Workday is charging per-credit for integration access to your own data in your own tenant. Every agent action. Every data query. Every API call. Metered. Priced. Tolled. It's not a huge line item on a per-call basis, but it's there.

That is not the pricing behavior of a company that believes it’s going to win the experience layer on its merits. If Workday genuinely believed Sana was going to be the front door that every employee uses every day, the agent actions would be free (or functionally free) and they’d monetize through stickiness, data gravity, and expansion into the adjacent systems. That’s the platform playbook. That’s what a company does when it thinks it’s going to own the next decade.

Flex Credits is the opposite playbook. It’s the pricing behavior of an incumbent that has already privately accepted utility status and is trying to maximize the current quarter before the market catches up. It’s extraction. It’s rent collection on the path to data that the customer already owns.

The “moat is a tollbooth” metaphor used to be about UX friction as a tax on users. Flex Credits makes it literal. It is a financial tollbooth on the drawbridge between your data and the experience layer you’re trying to build. Same metaphor. Two layers of meaning.

And here’s the thing about tollbooths. Sophisticated drivers find the back road.

The move this quarter: define your experience layer independently of whatever your ERP vendor is selling you.

Write the brief. What does your front door actually need to do? Who uses it? What does “good” look like? What security posture do you need? What integrations does it require into the fabric?

Answer those questions before a single vendor shows up at your door. Because if you show up to the Sana pitch without your own brief, you’re not evaluating a product. You’re watching a movie. And you’ll leave thinking you need a front door because you saw one on the screen, whether or not it’s the right door for your house.

(If it is Sana, great. Buy it. But buy it because it matched your brief, not because Workday already had your checkbook. Although, a compelling counter argument - Workday acquisition API calls are free. More to come on that Friday.)

The second move, parallel to the first: start building the skill on your team. Someone needs to own conversational design. Someone needs to own agent governance. Someone needs to own the UX review for anything that touches your employees. That is not an IT job. It’s a product job. And if the only product thinking on your HR tech team is “the business asked for it so we built it,” you are going to get steamrolled the minute the experience layer becomes the ballgame.

Tune the Instruments Before You Hire the Conductor

Three layers, three moves. Engine audit. Fabric inventory. Experience brief.

You’re doing them in parallel, not in sequence. But they have to be in proportion.

If your engine is on fire, don’t write the experience brief yet. If your fabric is the Invisible Spaghetti, an experience layer on top is just going to make the spaghetti invisible at higher resolution. The layers build on each other, and skipping the hard ones to get to the fun one is how you end up with a million-dollar AI agent politely surfacing broken data to the CHRO.

You cannot decouple what isn’t inventoried. You cannot orchestrate what isn’t governed. You cannot build an experience layer on a fabric that nobody can draw on a whiteboard.

Tune the instruments first.

The Point of All of This

Some very large organizations have figured this out. They built the engine-plus-experience-layer architecture when it cost tens of millions of dollars and required a platform team the size of a small bank.

The rest of us get to build it in 2026, when the same architecture costs a defined project budget, lots of AI tokens, and an experience brief.

The PeopleSoft comparison wasn’t a prediction of Workday’s death. It was a warning about what happens when one vendor confuses “the load-bearing wall” with “the whole house.” The practitioners who win the next decade aren’t the ones who leave Workday. They’re the ones who figure out how to make Workday disappear, while it quietly does its job underneath.

The goal isn’t to get off Workday.

The goal is to get your employees off Workday.

While it stays exactly where it belongs. Downstairs. In the basement. Running the plumbing. Out of sight.

P.S. Justin and I were having this exact conversation the other day. Not the architecture one. The router one.

Justin pointed at the mesh router (now placed on his workbench instead of being shoved in the rafters) in the basement and asked, “What is that box?”

Me: “That’s the internet.”

Justin: “So the internet’s really good down here because of that box, right?”

Me: “Yes. You are correct.”

That is the whole post. In two sentences. From a 12-year-old.

He didn’t care what brand the box was. He didn’t want to log into it. He didn’t need to understand the firmware. What he understood, instantly, was that the box makes the WiFi work, and the WiFi is what he actually cares about. Engine layer. Experience layer. Separated. Correctly.

If a 12-year-old can figure out decoupled architecture in one sentence, your HRIS team can too.

Make the plumbing invisible. The kids already figured it out.

— Mike

Director HR Tech | Basement Architect

The Department of First Things First. For the people who do the work.

tl;dr

• Workday won the last war (UX, cloud, APIs) the same way PeopleSoft won the one before it. Twenty years later, the pattern is repeating, with Workday in the PeopleSoft seat.

• The SOR isn’t dying. It’s being demoted to substrate. Value is moving up the stack to the system of insight / system of reasoning. Leadership has stopped asking “what’s our headcount” and started asking “what should we do about attrition in engineering.” Substrate is not useless. It’s deterministic. It’s foundation.

• Visier shipped an Analytic AI Agent Platform in December 2024. UKG calls Bryte “a system of action.” Eightfold just published “The End of the ATS Era.” These aren’t theories. They are shipping products.

• Workday is spending $1 billion-plus trying to catch up via acquisition — HiredScore, Evisort, Paradox, Flowise, Pipedream, Sana. Six deals in eighteen months. In 2023, their own VP of People Analytics publicly named the exact gap they are now trying to close.

• If you work inside a Workday tenant all day, you need to know what layer of the stack you are working on. It is not the one with the gravity.

The Founding Mirror

January 2005. Pleasanton, California. Dave Duffield is standing in a conference room inside the company he spent eighteen years building. It hasn’t, technically, been his for several days. The Oracle deal closed on January 7th. But there are still employees in that room who built their careers around the idea that PeopleSoft was different. They trusted him. He has to tell them it’s over.

“There were people crying,” he would later recall. “Men crying, in that room.”

He calls it the worst moment of his life.

Within weeks, Duffield and Aneel Bhusri walk out of Pleasanton and start over. Cloud-native this time. Same founding premise Duffield had in 1987: enterprise software shouldn’t be miserable to use. The competition back then was IBM dinosaurs; mainframe relics designed for administrators, not humans. PeopleSoft beat them because Duffield thought software should look like it was made for the people who had to use it.

They called the new company Workday.

Twenty years later, Workday is Pleasanton. It is the incumbent. It’s the system everybody else is building around. The UI that was revolutionary in 2006 is the UI that makes a retail manager want to throw a laptop out the window in 2026. The APIs that were once “open” are now table stakes, and in some cases a liability (more on that in a minute). The automatic updates are a regression, not a feature.

And somebody inside the building has been telling them what’s coming for at least two years. The call is coming from inside the house.

This is a piece about that.

The Demotion

Here’s the framing:

HR tech has had four eras. Each one has been disrupted by the next. We are inside the third disruption.

Era 1: the Mainframe. IBM. Unisys. The dinosaurs. Character-based green screens, transaction codes you had to memorize, batch payroll running at 2 AM. One HRIS admin for every 5,000 employees, a three-ring procedure manual on the desk, and a weekly phone call to the data processing department because nobody else in HR could talk to the system. Won on: centralization. The first time employee data existed in digital form at all. Died when client-server computing made distributed compute possible (and a former IBM engineer named Dave Duffield decided enterprise HR software shouldn’t be miserable to use).

Era 2: the On-Prem System of Record. PeopleSoft. SAP R/3. Oracle HCM before the cloud. Client-server architecture. SQL databases. Graphical interfaces. Real-time reporting. Help documentation that actually worked. Won on: dragging HR software out of the mainframe and into something that looked like the rest of the enterprise application stack. Died when the cloud rewrote the economics, SaaS rewrote the purchasing model, and mobile made every on-prem interaction feel like a time warp.

Era 3: the Cloud System of Record. Workday. SuccessFactors. Oracle HCM Cloud. UKG Pro. Dayforce. Same file cabinet, better UX, open APIs, SaaS economics, predictable releases. Won on: UI revolution plus integration openness. This was the game for a decade-plus. Workday won it, hard.

Era 4: the System of Insight. The System of Reasoning, depending on who you ask. Visier. UKG Bryte. Eightfold. Gloat. Phenom. Fuel50. Beamery. Every insight-first vendor is adding the ability to act. Every SOR vendor is bolting on insight. They are converging on the same position from opposite directions. Only one side has the history of disrupting the other.

The shift is subtle, but it is the whole ballgame. Leadership’s question has changed.

They do not ask “what’s our headcount.” Workday can answer that. They do not even ask “why is headcount growing.” Workday can help with that, if you spend a year cleaning your data model.

They ask: “What should we do about attrition in engineering?”

That is a fundamentally different product category. The SOR can’t answer it. The SOR can support an answer. It cannot produce one. And leadership has lost interest in systems that merely store what you already know. They want systems that tell them what to do.

The System of Record is not dying. Nobody is turning off payroll. Workday still handles the deterministic plumbing: I-9s, tax withholding, comp cycles, the stuff that keeps the lawyers off your neck. That work continues. That work is important.

It just isn’t where the value is anymore.

The System of Record is being demoted to substrate.

Who’s Actually Building It

If you want to know where the compass is pointing, watch what’s shipping.

Visier launched their Analytic AI Agent Platform in December 2024. Claimed to be the first of its kind in the category. They built agents that analyze people data and make recommendations, across whatever data sources you point them at, with full security and governance. Their CEO, Ryan Wong, said the quiet part out loud at the time: “Unlike other agentic platforms that focus on automation of workflows, we’re fixated on unleashing AI agents to analyze people and work data to quickly deliver the answers and recommendations that drive workforce impact.”

Then in November 2025, they shipped an MCP server. A universal connector letting any AI agent (Claude, Copilot, whatever) pull governed people insights with a single secure protocol. Ike Bennion, their VP of Product, said this on the launch call:

“The market is quickly realizing that building DIY agents on top of context-poor APIs is a recipe for chaos.”

Read that again with your Workday hat on. Whose APIs do you think he’s describing?

UKG Bryte, their agentic AI platform, is sold in UKG’s own press releases as “a system of action.” They are not hiding the framing. Jason Averbook (former HR digital transformation lead at Leapgen, now at Mercer) called the Workforce Intelligence Hub (Bryte’s operational command center) “a monumental leap forward... elevating workforce data to the same strategic importance as CRM and ERP.” An HCM vendor is publicly positioning its insight layer at the same altitude as Salesforce and Oracle ERP. Five years ago that would have been a pitch deck dream. Now it is a launch tagline.

Eightfold just published a blog post titled “The End of the ATS Era.” It opens with the sentence: “The ATS was built for record-keeping. The Intelligence Revolution demands action.” That is a vendor making the full-frontal attack on an entire category. You can call it marketing. You can also notice that they are saying it loudly and that nobody at Workday is disagreeing in public.

Around those three, the supporting cast tells you this is a pattern, not an outlier. Gloat on internal talent marketplaces (production-grade integrations with Workday and SAP). Phenom on the experience layer, which regular readers know I have been writing about for six months. Beamery on workforce planning fused with external labor market data. Fuel50 on career pathing, recently a Visier MCP partner. Galileo, Josh Bersin’s own AI tool, with 400-plus sub-agents integrated into ServiceNow Now Assist and HiBob.

Insight-first vendors adding agency. SOR-first vendors bolting on insight. They are meeting in the middle.

Only one side has the history of being the disruptor.

What Workday Is Doing About It (And Why It Rhymes With 2023)

Workday is not sitting still. Give them that.

They have:

• Acquired HiredScore (AI talent orchestration)

• Acquired Evisort (contract intelligence)

• Laid off 1,750 employees — 8.5% of the workforce — in February 2025 to “invest in AI”

• Acquired Flowise (low-code agent builder)

• Acquired Paradox (conversational AI for high-volume hiring)

• Acquired Sana for $1.1 billion (AI knowledge, learning, and agent platform)

• Acquired Pipedream (iPaaS with 3,000+ connectors)

• Replaced the CEO. Eschenbach out. Founder Aneel Bhusri back in, February 2026

• Lost their CTO, Peter Bailis (the face of Workday Illuminate, the AI strategy’s public front) to Anthropic. As a Member of Technical Staff. After eleven months in the job.

Six major acquisitions in eighteen months is not a strategy. It is a shopping spree with a corporate card and a deadline.

The Pipedream acquisition is worth flagging separately. Most coverage lumped it in with the rest of the AI buys, but for architects it’s the quietest and most significant signal in the sprint. Workday’s integration stack (EIB, Core Connectors, Studio) was built for a specific world: deterministic, enterprise-to-enterprise data flow on scheduled runs. That stack is good at what it was designed for. It is not the stack the agent era needs. Agents touch three thousand SaaS tools in real time, not ten ERP systems on a nightly batch. Buying Pipedream is a quiet acknowledgment that the integration model Workday built the company on isn’t the one they need to win the next war. That’s an architectural concession, and it’s the right one. It is also not something you say out loud at Rising.

The theory of what they are building is coherent enough. Sana as the new “front door for work,” replacing Illuminate as the AI brand. Flowise as the agent builder. Pipedream as the orchestration layer. An Agent System of Record on top to govern the whole fleet. Altogether: Workday becomes the insight layer and stays the system of record. One vendor, whole stack, all the budget.

I want that bet to pay off. I work in this ecosystem every day. A successful Workday pivot is good for me, good for my org, and good for anyone reading this who’s chosen to specialize in the platform. I am not rooting for Pleasanton to fail.

I am, however, noticing what the script looks like.

Because in 2015, Josh Bersin (the most widely-read HR tech analyst in the world) wrote, explicitly, that “Workday understands this from their PeopleSoft playbook.” The founding DNA of the company was “don’t let happen to us what happened to PeopleSoft.” That awareness is now the company’s tell.

And in 2023, Phil Willburn (Workday’s own VP of People Analytics, founding member of the function since 2017, former US Intelligence Community analyst, a practitioner running people analytics for Workday’s own 18,000 employees with Workday’s own tools) told David Green on the Digital HR Leaders podcast, on the record:

That sentence went out to podcast subscribers in 2023.

That was before Visier built the Analytic AI Agent Platform. Before UKG shipped the Workforce Intelligence Hub. Before Eightfold declared the end of the ATS era. Before Workday itself started writing checks to close the gap Phil had already named in plain English on somebody else’s podcast feed.

A senior Workday insider diagnosed the exact problem the insight-layer vendors are now eating Workday’s lunch on. He did it publicly. He did it two-plus years before the acquisition sprint started. Product didn’t pick up. Or picked up too slowly.

This is what the innovator’s dilemma looks like rendered on a specific timeline with a specific quotable source. Workday knew. The internal analysis was there. The diagnosis was published. And the organization (twenty thousand people, $9 billion in revenue, a Gartner leader every year) couldn’t move fast enough to act on it until the threat was already shipping.

PeopleSoft had the same problem in 2003. Too many fronts. Too many acquisitions to integrate. An existing revenue stream they couldn’t cannibalize. Duffield saw it. Couldn’t stop it.

Oracle paid $10.3 billion.

I am not predicting that outcome for Workday. I am noting that the script they are running off is one we’ve seen before, that the previous run ended in a Pleasanton conference room with grown men crying, and that the guy whose name was on the building saw it coming and couldn’t stop it either.

But What About Extend?

I can hear the seasoned architects already. “You can’t argue Workday is losing the action layer without addressing Extend.”

Fair. Let me address it.

Extend is Workday’s billion-dollar bet that you’ll build your custom apps inside the walled garden. Inherit the data model. Inherit the security. Stay in the family. It’s pitched as the moat that keeps the action and reasoning layers tethered to the platform.

It’s also Workday’s PeopleTools moment.

PeopleSoft had PeopleTools. Same pitch, same architectural promise, same lock-in story. Customers built on it. They built deeply. And those PeopleTools investments became the technical debt that made the cloud migration brutal. The customization platform didn’t save PeopleSoft. It accelerated the dependency and made the inevitable platform shift hurt more on the way out.

Extend is running the same playbook. Adoption is concentrated in a thin slice of enterprise customers with real engineering teams (the rest of the install base has no Extend footprint and isn’t going to develop one). The credit economics make build-vs-buy math favor buy in almost every scenario. External vendors ship in weeks; Extend builds take months. And if Extend was actually keeping the experience layer inside the garden, Workday wouldn’t have spent a billion dollars on Sana to bolt one on top.

Extend doesn’t keep the value layer inside the walled garden. It builds a porch outside the back door. And the customers most invested in Extend become the most locked into a substrate they can’t easily leave.

That’s not the escape from PeopleSoft. That’s the same script.

What This Means for You

If you spend your days inside a Workday tenant (building business processes, maintaining integrations, writing calc fields, fighting with security groups, shipping BIRT reports nobody will read) you are working on the substrate.

That is stable work. That is real work. That is work that will continue to exist for as long as payroll has to be deterministic, which, legally, is forever. I am not telling you to quit. I am telling you to know what layer you are on.

Because the value layer - the part that gets budget, the part that gets keynotes, the part that turns into career-defining projects - is moving up.

If you want to work there, build the skills that live there. Data modeling is still table stakes (your SOR skills translate cleanly; you will be shocked how much of a Workday Prism build maps directly to an insight-layer product). But the new skill is reasoning about what the data means and who is allowed to act on it. Insight delivery as a product discipline. Agent governance frameworks. Knowing the difference between an agent that recommends and an agent that acts, and designing the guardrails between them.

And here’s the part regular readers have heard from me before: the insight layer is only as smart as the substrate beneath it. A reasoning agent fed garbage data produces garbage recommendations, just with more confidence and a prettier dashboard. Which means the Workday admin who keeps the plumbing clean and thinks about what the reasoning layer needs from it becomes more valuable in Era 4, not less. The career risk isn’t being on the substrate. The career risk is being on the substrate and only thinking about the substrate.

Watch Visier. Watch UKG Bryte. Watch Gloat. Watch Phenom. Not because you have to leave Workday. Because in five to ten years these vendors will either be the next incumbents or the acquisition targets of the current ones, and in either case the people who know how to architect on top of them will be in the rooms where the decisions get made. The Workday-only specialist in 2026 is in a similar spot to the PeopleSoft-only specialist in 2008. Not unemployed. Not underwater. Just working on a layer that fewer and fewer people are talking about.

Most practitioners don’t see this shift yet. Some do and aren’t talking about it. A few are quietly updating their résumés.

The ones who will come out of the next five years with the best careers are the ones who understand, right now, which layer they are building on, and are honest with themselves about whether they want to stay.

Close

PeopleSoft is still around.

Oracle still bills for it. The product still works. Public sector uses it. Universities run it. It does the things it has always done. You just don’t hear about it anymore. That is not death. That is the demotion. That is what being the substrate looks like after somebody else moves up the stack. It is also not where HR tech careers go to grow. Substrate pays salaries. It does not attract new ones.

Workday might not go there. I am not writing PeopleSoft’s obituary a second time. Sana is a real product. Flowise is a substantive acquisition. Pipedream gives them something Workday has genuinely never had; a living integration ecosystem with a developer community attached. Bhusri coming back to run the company is founder-mode energy and that is not nothing. There is a version of the next five years where Workday successfully buys its way from Era 3 to Era 4 and keeps the whole enterprise locked in through the transition.

There is also the other version.

The last time a company with twenty thousand employees tried to defend its substrate by acquiring everything around it, the guy in the Hawaiian shirt had to tell crying employees in a Pleasanton conference room that it was over.

And before he had to do that, somebody inside the building had probably been telling him for years.

P.S.

Justin asked if he could have my old MacBookPro that sits in the basement, still plugged in, taped power cord and all. I told him no, that one runs the media server. He asked what that meant. I explained that the 2008 MacBook Pro downstairs has been quietly serving movies and music to every device in the house for years, and will keep doing it until it stops, at which point I’ll probably finally go fully cloud (we’re like 90% of the way there anyway). He thought about it for a second and said:

“So it still does a job. It just doesn’t get to come upstairs anymore.”

Yeah, bud. That is what a System of Record looks like after it gets demoted. Still on. Still running. Still doing real work. Just doesn’t get to come upstairs anymore.

— Mike

Director HR Tech | Hawaiian Shirt Adjacent

The Department of First Things First. For the people who do the work.

You just watched me talk for ~20 minutes about RAG, MCP, agents, and a 12-year-old who called the future of HR AI "the Wikipedia of work." If you scanned the QR code, you're already here, so I'll spare you the bio you just sat through.

Here's what this place is:

The Department of First Things First is a weekly-ish newsletter for the people who actually have to make Workday work. Not analysts. Not buyers. Not the conference circuit. The people who go back to their desks on Monday and configure the business process, audit the security model, and explain to a CFO why the agent demo isn't shipping in their tenant by Q3.

The talk you just sat through is what this newsletter sounds like every week. Same voice. Same lens. Same anti-encyclopedia stance.

Applied to whatever the ecosystem is doing this week.

Here are five posts to tell you whether to subscribe:

1.

Workday Didn’t “Fix” Their UI. They Just Bought a $1 Billion Skin.

Michael Domingo

Jan 28

Justin came to me yesterday asking for $10 to buy a “skin” for his video game.

Read full story

My son Justin asked me for $10 to buy a cosmetic skin for his video game character. It didn’t make him faster or stronger. He just looked cool while falling off a cliff. Then Workday released their Sana-driven UI and I realized they’d done the exact same thing for a billion dollars. This post breaks down what the new UI actually means for practitioners, where the security nightmares hide, and why a search bar on top of a messy warehouse doesn’t clean the warehouse.

2.

April 27

I promised I’d share this during the talk. Workday rebranded their entire AI strategy in nine months. Again. Here's what's actually shipping under the Sana banner, what's still slideware, and why the Illuminate-to-Sana pivot doesn't change a thing about what you need to build in your tenant.

3.

The Moat is a Tollbooth: Why the AI Panic Around ERPs is Only Half Wrong

Michael Domingo

Feb 25

tldr; Josh Bersin published a piece today arguing the market is irrationally punishing enterprise software (like Workday) over AI fears.

Read full story

Bersin argued that enterprise software is protected by a moat of decades-deep business rules. He’s right about the moat. But for the people actually using the system, that moat feels like a ten-lane tollbooth. This post introduced the framework I keep coming back to: the ERP isn’t going away. But the future is the ERP becoming headless infrastructure, buried so deep under experience layers and AI agents that the end-user forgets it exists.

4.

Put Down the Wrench: The Psychological Shift of Becoming a Solutions Architect

Michael Domingo

Feb 23

tldr; Transitioning from a Workday builder to an Architect isn’t about learning new modules. It’s a brutal psychological shift from fixing technical tickets to navigating corporate politics.

Read full story

The hardest part of moving from Workday analyst to architect isn’t learning a new module. It’s the psychological shift from fixing tickets to navigating politics. I break down what that transition actually feels like: the muscle memory trap, the instant gratification desert, and why the Kitchen Table Test (can you explain your architecture to an 11-year-old?) is the only metric that matters.

5.

I Let My 11-Year-Old Audit Workday’s UI. He Lasted 3 Minutes.

Michael Domingo

Feb 9

We talk a lot about “User Adoption” in HR Tech.

Read full story

I Let My 11-Year-Old Audit Workday’s UI. He Lasted 3 Minutes.

I handed my son Justin a Workday tenant and asked him to request a day off. He’s a digital native who can navigate any app in seconds. He lasted three minutes before giving up. If a kid who grew up with an iPad needs a Quick Reference Guide to use your system, your design is broken. This post is the one that makes non-technical leaders finally understand the UX problem.

If any of those landed, you’re in the right place. Subscribe and I’ll show up in your inbox weekly with architecture takes, vendor accountability, and the occasional story about my son accidentally explaining enterprise strategy better than a $50,000 consultant.

— Mike

Director HR Tech | Keeper of the Layer Cake

P.S. If you want to hear me talk instead of read me type, I’ve been a guest on:

• Talent Experience Live with Devin Foster at Phenom — “Speed Meets Experience” (to be published)

• Unsafe Harbor with Christian Delcid — Episode 14

• Workday Gold with Keith Bitikofer — December 19th, 2025

All are worth your commute.

tl;dr: Gartner's Trend #1 for 2026 is AI-Native Development, and the industry buzzword is "Vibe Coding." In Workday-land, that means business analysts prompting apps into existence via Extend and the AI App Builder. The apps work on Tuesday. They break three weeks later, and nobody on the team can debug what the AI wrote. Architecture is the why. Vibes are the what. You need both, in that order.

My son Justin is now 12 🥳, and he loves AI image generators (especially Meshy). He'll type in a prompt for a "cool space robot," and the model produces something that looks awesome at first glance.

Then he looks closer.

The robot has seven fingers. Its right foot is actually another left foot. The lightsaber is a flashlight with a ribbon on it.

Justin wants to fix it. But he can't actually draw. Not a robot, anyway. He only knows how to regenerate. So he hits the button and hopes the next version is better. Usually it just has different problems. A missing arm. Eyes pointing in opposite directions. A dragon with three nostrils (don't ask).

Keep that image in your head for a minute, because Gartner just put it on the front page of its 2026 Top Strategic Technology Trends.

Enter the Vibe

Gartner's Trend #1 for 2026 is AI-Native Development, and the buzzword the analysts are rallying around is "Vibe Coding."

The pitch: generative AI has gotten so good at producing software that we don't need to write code anymore. You describe what you want, and the machine vibes it into existence. No syntax. No semicolons. No deep technical knowledge required.

Analysts love this. It's clean, it's democratizing, it fits on a slide. And in the Workday ecosystem specifically, it's about to cause a massive architectural mess.

The Build-vs-Buy Plot Twist

Remember when we all agreed we bought Workday specifically so we wouldn't have to build anything? Yeah. Gartner thinks that's over.

The prediction: by 2030, 40% of enterprise application portfolios will be custom-built on AI-native platforms, shifting the build-vs-buy equation back toward building. In Workday-land, that looks like a Tuesday.

It's a Tuesday. A smart, ambitious business analyst, fluent in PowerPoint and light on XpressO, opens Workday Extend and the new AI App Builder. Her VP wants a "quick app" to track something the core product doesn't handle. A specialized leave type, maybe. Or a contingent worker checklist. Or a talent mobility micro-workflow/survey.

She types a prompt. The App Builder thinks for a minute. Then, magic. An app appears. Fields. Business process. A condition rule. A notification. She clicks through it. It works.

She demos it Thursday. Her VP is thrilled. Her VP's VP is thrilled. Someone uses "citizen developer" unironically. She gets a spot bonus.

Three weeks later, the hyperdrive engine falls off the space robot.

The Seven-Fingered Robot in Your Tenant

The app breaks because somebody rolled out a new supervisory org the condition rule didn't understand (because it hardcoded the org list). Or a condition rule upstream changed. Or the custom object's reference id got renamed during a data cleanup (this one makes me crazy). Or the vibed calculated field worked in the sandbox but stopped working after the next weekly service update.

Pick one. Doesn't matter. Something always changes in a Workday tenant. That's the whole job.

Here's the problem. The BA didn't write the app. She prompted it. She doesn't know which calculated field is doing the lookup. She doesn't know why the condition rule fires in that specific order. She doesn't know what the custom object's security profile inherits from. She can't debug it, because there's nothing in her head to debug from. The code doesn't exist in her brain. It exists only on the screen.

So she does the only thing she can do. She goes back to the App Builder and types: "Fix the app."

The App Builder cheerfully produces new code. Sometimes it works. Sometimes it breaks something else. Sometimes it silently changes the behavior in a way nobody notices until the quarterly audit.

Regenerate. Hope. Repeat.

That's Justin with the robot. Except now Justin is in production, and the robot has access to your worker data.

Extend Isn't a Toy

Here's the part I want every HR tech leader reading this to seriously sit with for a second:

Workday Extend is not a low-code platform with an AI wrapper. It's a full development environment that happens to have an AI wrapper.

That distinction matters. Low-code platforms have guardrails. They constrain what you can do so the blast radius stays small. Extend doesn't work that way. Extend gives you access to the core Workday object model. Custom objects. Calculated fields. Business processes. Integration endpoints. Security groups.

If you vibe a website, the CSS looks weird. If you vibe a Workday app, you might accidentally create a security group that leaks your CEO's home address.

And it gets worse. Gartner also predicts that 80% of organizations will evolve their engineering teams into smaller, AI-augmented teams by 2030. Translation: fewer humans, more vibes. Which sounds efficient until the BA who prompted the app leaves the company, and the three admins remaining on the team have never seen the code, never seen the spec, and can't tell you why the condition rule fires the way it does.

That's not a tiny team. That's an accountability gap with a headcount rationalization on top.

What to Do (Because I'm Not a Luddite)

I'm not saying don't use the AI App Builder. Do it. I’m doing it too (Want proof? Check my GitHub). I'm saying use it the way senior developers use GitHub Copilot: as a drafting tool, not an authoring tool. The draft is free. The review is the job.

Gartner's own 2026 action plan gets this mostly right. Three moves worth stealing:

Establish a platform team. Centralized oversight. Standards. Governance. Somebody who can say "no, that's not how we do it here" without getting called change-resistant.

Implement real guardrails. Code review. Compliance checks. Security profile audits. Don't let an app ship to production just because it passed the "does it run in the sandbox" test.

Adopt an AI-first mindset, not a vibes-only mindset. Accelerate delivery, yes. But only after your team understands the architecture the AI is generating on their behalf.

I'd add one more, Workday-specific: every vibed app needs a named human owner in your Extend inventory. When the app breaks during the next release cycle, you know who gets paged. "The AI built it" is not a name on a support ticket.

The Bottom Line

Vibe coding isn't democratization. It's delegation without accountability.

Architecture is the why. Vibes are the what. You need both, in that order, or you're just prompting your way into a postmortem.

Justin has figured out that the generator isn't going to get the robot right on its own. He's started asking me to help him fix the fingers.

Your business analysts will figure that out too. The only question is whether they do it before or after the app ships during Open Enrollment.

— Mike

tl;dr — Josh Bersin published a defense of Workday this week, responding to the a16z piece arguing Workday is structurally cooked. Buried six paragraphs from the bottom is a single sentence saying the pioneers of agentic HR will be customers and IT shops, not vendors. He is correct. He has also accidentally written the mission statement of this newsletter, three paragraphs above a plug for the AI layer he wants to sell us.

The line is buried, six paragraphs from the bottom of a 2,000-word post. I almost missed it.

That is the entire thesis of practitioner-led HR Tech, written in someone else's blog, by someone who has been doing this longer than I have.

It is exactly right.

It is also completely undermined by the other 1,900 words.

The Setup

For context. A partner at a16z (Joe Schmidt IV) published a long piece last week called "Workday's Last Workday?" The thesis is that HCM is the last large enterprise category without an AI-native challenger, and Workday's moat is lock-in disguised as love.

Buried in that piece is a sentence, attributed to a Workday admin Schmidt interviewed, describing Illuminate as "the same manual admin work with a chat surface bolted on."

That sentence could have been pulled from this newsletter. I have written some version of it at least five times.

Bersin's response this week is titled "A16Z Perspectives On Workday Seem Naive." He defends the consulting ecosystem. He likes the rules-as-rails framing. He says the future will arrive slowly. Some of which I agree with.

Then he buries the most important sentence in the entire HR Tech discourse this month at the bottom of the post and follows it with a Galileo plug.

The Contradiction

Most of the essay is a defense of vendor architecture. The deterministic engine. The Agent Passport. The Agent System of Record. The fact that Workday Rising has 20,000 happy attendees, which apparently proves customers do not really mind the product.

Then, near the end, Bersin takes a sharp turn.

He talks about how new architectures take years. How Workday itself was a thin platform that only became a $10B business after years of customer-led iteration. How the next wave of innovation will look more like "build it yourself" than "wait for the big vendor to deliver it all."

And then the line.

Which is correct. Which is the entire game. Which is why I write this newsletter.

But here is the question I cannot get past.

If practitioners are the pioneers, why is the rest of the post a defense of the vendor's worldview?

You cannot have it both ways. Either the pioneers are the customers (in which case the move is to arm them, not reassure them that the vendor's architecture is fine and slow innovation will protect everyone) or the vendor's architecture is the future (in which case practitioners are not pioneers, they are operators).

Bersin picks both. The post reads like a love letter to Workday with a footnote that says "by the way, the people in the trenches will figure this out."

(Generous title to bestow on the practitioners from inside the Sana launch press release.)

What Practitioner-Led Actually Looks Like

Here is what Bersin's last paragraph gets right, even if the rest of the post does not commit to it.

The teams I see making the most progress on Agentic HR right now are not waiting for Sana. They are not optimizing their Flex Credits spend. They are not standing in line for the Agent System of Record.

They are doing the unsexy work first.

Auditing job profiles that have not been touched since implementation. Documenting business processes nobody has looked at since 2019. Mapping their integration spaghetti into something resembling an inventory. Teaching their analysts to read an API spec.

Tuning the instruments before hiring the conductor.

When the vendor finally ships the agent, those teams will plug it into a clean system and see real output. Everyone else will plug a chat interface into a tenant full of garbage and wonder why the agent hallucinates.

That is what practitioner-led actually looks like. It is not glamorous. It does not require a billion-dollar acquisition. It does not get a press release.

It requires somebody to do the work nobody wanted to do.

The Bottom Line

Bersin's last sentence is correct. The pioneers will be the customers and IT shops. The people in the tenant. The people who get the 2 AM call when payroll breaks. The people who own the data model the agent eventually has to live inside of.

That is the whole thing. That is what this newsletter is about. That is why I write it.

The fact that the most accurate sentence in the most prominent HR analyst post of the week is one paragraph long, buried at the bottom, and three paragraphs from a Galileo plug, is its own kind of signal.

The work is ours. The credit, eventually, will be too.

We just have to do it before the analysts notice.

— Mike

Director of the Department of First Things First | Newly Crowned Pioneer

The Department of First Things First. For the people who do the work.

P.S. Justin asked what I was writing about. I told him an analyst said the people who actually do the work are going to be the heroes of this story.

He thought about it. "Dad, that's like when I do the group project and Skyler tells everyone what we did."

He is twelve. He gets it.

a16z partner Joe Schmidt just published a piece arguing HCM is the last enterprise category without a serious AI-native challenger, and they want to fund the company that replaces Workday. He even put his email in the post.

The diagnosis is half-right. The UI is brutal. Admins do absurd workarounds. Illuminate is a chat surface bolted on top of a forms-and-approvals engine. All true.

The cure is a fantasy. The reason you can't replace Workday isn't its architecture. It's that your tenant is a ten-year archaeological dig of decisions nobody documented.

A new vendor doesn't dissolve that fortress. You do. Again. From scratch. For three more years.

Joe Schmidt at a16z published a piece this week called "Workday's Last Workday."

I want to be clear up front: it's good. It's not the lazy "AI will eat ERP" take we've been seeing on LinkedIn for two years. He talked to admins. He understands the Illuminate critique. And he literally put his email in the post: build this, I'll fund you.

Respect.

But this is where the venture brain and the practitioner brain part ways.

Schmidt thinks the moat is the architecture. He thinks 100% renewals signal lock-in, not love (correct), and that the lock-in is technical (incorrect). He thinks if you build a multi-tenant HRIS in 2026 with agents instead of forms, the Fortune 500 rips out Workday the way they ripped out PeopleSoft.

They won't. And not because Workday is invincible.

Because the fortress isn't Workday.

What your tenant actually contains

When you log into your Workday tenant, you are not looking at "an HRIS." You are looking at a decade of decisions your company made about how it runs.

Why does Job Profile 47 require an extra approval from Legal? Because in 2019 someone in Compliance flagged a weird thing during an acquisition and a step got bolted on. Nobody wrote it down. Janet remembered. Janet retired in 2023.

Why does the comp business process branch differently for grade 19 versus grade 20? Because the original design decision was made in a conference room in Pleasanton during implementation, the consultant left, the BP still works, and nobody is going to touch it.

Why are there four custom fields on Position called "Strategic_Indicator_Old_DO_NOT_USE"? Because we tried to deprecate them in 2021, three integrations broke, we ran out of cycles, and now they're cursed.

Multiply that by every business process, every security group, every condition rule, every calc field, every validation, every integration. That is your tenant.

That is the fortress.

Schmidt's pitch is: rebuild this on agents instead of forms.

Cool. Who's going to make the four thousand decisions about how it should work?

The same people who made them the first time. Sitting in the same conference rooms. With the same competing priorities. For three more years.

The Tessera analogy doesn't hold

Schmidt points to Tessera doing AI-native SAP migrations at Fortune 500 scale and argues HCM is the same shape of problem at smaller surface area.

It isn't.

ECC-to-S/4HANA is an SAP-to-SAP migration. Vendor-supplied tooling. Vendor-supplied schema. Vendor-supplied incentives (SAP literally end-of-life'd the old version). Workday-to-Greenfield has none of that. There's no shared schema, no migration toolkit, no friendly Workday SE on the call helping you map fields. Every business object has to be re-derived from scratch by humans who already have day jobs.

You're not migrating data. You're re-deriving institutional knowledge.

The seventeen comp adjustments

Schmidt opens with a great image: an HR admin copying seventeen comp adjustments from a spreadsheet into Workday, one field at a time, while a business partner watches on Zoom.

Yes. That happens. I have been in that meeting.

But ask the next question. Why is the admin doing it that way?

It's not because the comp module can't bulk-load. It can. It's because someone, somewhere, doesn't trust the data feeding it. Or there's a legacy approval workflow that doesn't fire correctly when you push it through the API. Or Compliance inserted a manual review step in 2018 and nobody has the political capital to remove it.

The AI agent answer is: automate the workaround.

The architect answer is: kill the workaround.

These are not the same thing. And the AI-native challenger that ships in 2027 will be automating workarounds in their own tenants by 2030. Because that's what humans do. We accumulate workarounds. The system does not.

What $40M actually buys

If someone takes Schmidt's email seriously and builds this thing, here's what they'll discover after three years:

They will have rebuilt Workday. Not "Workday for the AI era." Workday. With a chat interface. Because the Fortune 500 will demand the same compliance, the same payroll determinism, the same security model, the same audit trail, the same data residency, the same SOC 2, the same FedRAMP. And over five years the new vendor will accumulate the same forms-and-approvals scar tissue, because the customers will demand it.

The new system of record looks exactly like the old system of record. Only the founders are richer.

What's actually happening

The future of HR Tech isn't a Workday replacement. It's a Workday subordinate.

The deterministic engine stays. Payroll runs. Compliance holds. The system of record does what it is good at: keeping the company out of jail and the paychecks correct. Sitting on top of that, the experience layer goes feral. Conversational interfaces. Agents that handle orchestration. Sana, Phenom, ServiceNow, whatever wins the front door. You never see the BP. You never click through fourteen steps.

That is happening. It is happening now. It does not require a16z to fund a greenfield.

It requires the practitioners who actually run these systems to do the unsexy work of cleaning the data, rationalizing the integrations, and building the governance layer that makes the agents trustworthy.

The fortress isn't Workday.

The fortress is the body of decisions encoded in your tenant.

A new vendor doesn't dissolve that fortress.

You do. With a wrench. On a Tuesday morning.

— Mike

Director HR Tech | Tenant Archaeologist

The Department of First Things First. For the people who do the work.

P.S. Justin came to me last week and said he could "build a better Roblox in a weekend." I asked him what game engine he would use. He said "what's a game engine." Dear reader, a16z would fund him.

If you’ve been trying to keep up with Workday’s agent announcements over the past twelve months, I have some bad news and some good news.

Bad news: you’re going to need to learn a bunch of new product names. The “Illuminate” branding that Workday rolled out in 2024 is now mostly gone, replaced by “Sana from Workday” following the Sana acquisition that closed in November 2025. A bunch of agents that you might have read about under the Illuminate banner (”Illuminate Recruiting Agent,” for example) are now just “Recruiting Agent” inside the Sana umbrella.

Good news: the architecture underneath hasn’t really changed. The agents are still grounded in Workday data, still governed by the Agent System of Record, still inherit tenant security. The branding got a haircut. The plumbing is the same plumbing.

I’ve been saying this for a while now, and I’ll keep saying it: vendor AI branding has a shelf life of about nine months. The architecture is permanent. If you’re building your HR tech strategy around product names, you’re going to rebuild it three times a year. If you’re building it around data, governance, and workflow, you’re going to be fine.

With that framing, here’s the practitioner’s field guide to the current (April 2026) Workday agent lineup.

Before You Touch Anything: The UMSA

One thing Keith Bitikofer covered in his Philly RUG session that deserves to be front-loaded: if you want to use Workday Agents, you need to be on the UMSA. This replaces the old Innovation Services Agreement. It’s the gate. If your org is still on the MSA without the UMSA addendum, you’re not turning on agents yet. Check with your CSM before you plan anything else.

Start With the Platform, Not the Agents

Before we get into the roster, this is the part most agent write-ups skip. And it’s the part that actually matters.

Workday Agent System of Record (ASOR) is the control plane. Every Workday-built agent, and increasingly every third-party agent that wants to play nicely, registers here. ASOR is where you manage what the agent can see, what it can do, who can invoke it, and what data it’s allowed to touch. It’s the governance layer for your entire agent fleet.

ASOR is also the attestation layer. It’s the ledger that records where each agent came from, what version of the logic is running, which skills are enabled, and which humans are accountable for it. “We don’t know what agents are running in our tenant” is how you end up in a regulatory filing. ASOR exists so that sentence never gets spoken out loud in your company.

You access ASOR through the Agent Management Hub, which includes an Agent Registry, a view for Unregistered Workday Agents (yes, that’s a real section, and yes, you should check it), and links out to the Agent Marketplace. Third-party agents from Microsoft, Azure, AWS, Google Cloud, and partner ecosystems can register into ASOR using shared protocols including MCP and Agent-to-Agent (A2A). ASOR is becoming the universal control plane for agents, whether they were built in Workday or not. That’s a much bigger story than any individual agent on the roster.

If your HR tech team is going to care about exactly one Workday AI thing this year, care about this: an unmanaged agent in your tenant is a risk. A managed agent with a clear skill inventory, audit trail, identity, and permission model is a tool. ASOR is what moves you from the first thing to the second. Justin figured this out the hard way when I gave him unsupervised access to YouTube. Same concept, much higher stakes.

Workday Build is the platform for creating your own custom AI solutions inside Workday. This is Workday’s answer to “we want to build our own agents too,” which, if you read my last post, you know I have a few feelings about. Build is more structured and involves real developer workflows. The same rule applies: if no human on your team understands what the custom agent actually does, you don’t have an asset. You have a liability with a friendly UI.

Sana Agent Builder Pro (formerly called Flowise Agent Builder, if you’ve seen that name floating around) is the no-code/low-code tool for creating custom agents. It’s Build’s friendlier cousin, aimed at configurators rather than developers.

Flex Credits is the pricing model for all of this. We’ll come back to this one in detail, because it deserves more than a sentence. It deserves its own section.

The Front Door

These two are the agents your employees will actually touch. Everything else happens behind the scenes.

Sana for Workday

What it is: the new AI-native interface for Workday itself. Ask it questions, run transactions, get reports, navigate workflows, all in natural language. Replaces a lot of the classic Workday UI for users who prefer to chat rather than click.

My take: this is the most important shift in the Workday user experience since the 2016 redesign. For power users, it’s a convenience. For casual users (managers, frontline employees, anyone who uses Workday three times a year and forgets where everything is), it’s the difference between getting something done and filing a ticket. If you’re on a Workday tenant today, your employees are going to expect this within 18 months. Plan the change management now.

Watch out for: the security model still applies. Sana can only surface what the user’s permissions allow. This is a feature, not a bug, but it means “why can’t Sana show me my team’s salaries?” is going to be a recurring support ticket (depending on your security setup. I've seen some doozies, including several where manager didn't have access to pay info!). Get your comms ready.

One more thing Keith flagged in his RUG session that stopped me cold: Sana user access is managed separately from Workday. Your existing Workday security model does not automatically carry over. If you’re expecting seamless permission inheritance, you’re going to be surprised. Budget time for a separate access management workstream before you roll this out.

Sana Self-Service Agent

What it is: the workhorse. Ships with 300+ skills out of the box, covering the standard HR and finance self-service requests: address changes, PTO balances, benefit elections, expense inquiries, all the greatest hits.

My take: this is the agent that will replace the most tickets, and therefore the agent with the clearest ROI story. It’s also the one your Tier 1 HR support team is going to have feelings about. Have that conversation early.

Watch out for: “300+ skills” sounds impressive. In practice, the skills vary in maturity. Test the ones specific to your workflows before assuming they all work the way the demo works. And per Keith: you can’t use Proxy to test Agents or Sana Core. If Proxy is part of your standard testing workflow (and for most admins it is), you’re going to need to rethink how you validate before you go live. Plan for that extra testing time now.

The Talent Stack

These are the agents aimed at the recruiting, mobility, and frontline parts of the employee lifecycle.

Recruiting Agent

What it is: AI assistance across the recruiting workflow. Sourcing, screening, candidate communication, interview scheduling, the usual suspects.

My take: this is where Workday’s Hiredscore acquisition comes home to roost. The agent is meaningfully better than Workday’s pre-Hiredscore recruiting AI was (umm, crappy keyword matching ain't AI, Mr. Mobley). If your TA team still thinks of Workday Recruiting as a glorified ATS, this is the year that assumption may actually break.

Watch out for: the boundary between this agent and Paradox (which Workday also owns via partnership) is not always obvious. Ask your CSM which use cases go where. Also: the pricing on this agent has teeth. We’ll get to it.

Talent Mobility Agent

What it is: personalized internal matching. Surfaces roles, gigs, and projects for employees based on skills, goals, and history. Helps managers and recruiters find internal candidates.

My take: the use case is great. The execution depends almost entirely on your skills data. If your tenant’s skills model is thin or inconsistent (and let’s be honest, whose isn’t), this agent is going to surface noise. Invest in the skills data first, turn on the agent second. Also: the pricing on this one has teeth too.

Candidate Sourcing Agent

What it is: proactive identification of contingent talent for open needs. Aimed at procurement-adjacent hiring.

My take: niche but valuable for companies with large contingent workforces. If your contingent spend is significant, worth a look. If not, skip it for now.

Frontline Agent

What it is: self-service and workflow automation purpose-built for frontline workers. Shift questions, schedule swaps, benefits queries, the stuff hourly employees actually need.

My take: the most undervalued agent on the list. Frontline workers have been the stepchild of HR tech for two decades. An agent that actually works on a phone, in a break room, in 90 seconds, is a real unlock. If you’re in retail, hospitality, healthcare, logistics, or manufacturing, this one is worth a hard look.

The Operations Layer

These are the agents that live in the back office. Your employees will probably never see them. Your ops teams will.

Payroll Agent

What it is: automates payroll processing tasks, improves data quality, reduces manual reconciliation.

My take: payroll is the most conservative corner of HR tech for very good reasons (if you get payroll wrong, everyone notices). This agent is not going to replace your payroll analyst. It might eliminate 20% of the boring work that currently fills their day. That’s the right bar.

Watch out for: don’t let anyone pitch this as a headcount play. It isn’t. It’s a quality and cycle-time play.

Business Process Optimize Agent

What it is: analyzes your tenant’s business processes and recommends improvements. Think of it as a Workday-native process mining tool with opinions.

My take: potentially the sleeper hit. Every Workday tenant I’ve ever seen has business processes that have been amended, patched, and condition-ruled into Rube Goldberg machines. An agent that can spot the redundant approval step, the orphaned notification, the condition rule nobody remembers writing, is doing real work.

Watch out for: recommendations are only as good as your willingness to act on them. If your governance process requires six signatures to change a BP, the agent is going to produce insights that die in committee.

Deployment Agent

What it is: an AI assistant specifically for Workday administrators and implementers. Ask it anything about Workday configuration, troubleshooting, calculated field logic, business objects, you name it. Sourced from the Admin Guide, User Guide, Community Accepted Answers, release notes, and configuration documentation.

My take: this one should have been higher in the article. For the people reading this newsletter, it’s the most immediately useful agent on the list. 120,000 questions answered in February 2026 alone, with 49% request growth from January to February. Those aren’t vanity metrics. That’s a signal that admins are actually using it.

The Workday-cited stat is a 20-25% time reduction in research and troubleshooting. Based on how much time I’ve personally watched admins spend digging through Community threads for an answer that should have been findable in five minutes, I believe it.

One more thing, and this matters: as of the 2026R1 release, the Deployment Agent is available in your Customer Central tenant, does not require the UMSA, and does not consume Flex Credits. It’s free to run today. Which is a good segue.

The Finance-Adjacent

I’m not going to dwell here because this newsletter is for HR tech practitioners, but for completeness: Contract Intelligence Agent, Contract Negotiation Agent, Financial Audit Agent, Planning Agent, Revenue Contract Agent, and Supplier Contract Agent all exist, are either GA or in early adopter status, and live on the Finance side of the house. If your role straddles HR and Finance, worth a conversation with your Finance counterparts about what they’re piloting. Also, hold onto the Contract Negotiation Agent name. We’re coming back to it.

The Cross-App Extension

Sana Enterprise

What it is: the piece that extends Sana agents beyond Workday itself. Connects to Gmail, Microsoft Outlook, Salesforce, SharePoint, and other enterprise systems. Lets the agents find, orchestrate, and complete work across applications, not just inside Workday.

My take: this is the most ambitious piece of the March 2026 announcement, and the one most likely to get messy in the wild. The pitch is seductive: one agent, all your systems, completes the work. The reality is that every enterprise integration has its own quirks, its own permission model, and its own “hmm, that used to work” failure mode.

If you’re going to pilot Sana Enterprise, pilot narrowly. Pick one or two systems, not five. Prove the pattern before scaling it.

Watch out for: data residency. When a user in Frankfurt asks Sana Enterprise to “summarize the last five emails about this candidate,” where exactly does that summarization happen, and across which legal jurisdictions does the data travel? Global HR tech leaders have spent a decade carefully building tenant architectures that respect EU, APAC, and Latin American data boundaries. Cross-app agents have the potential to undo that work in a single well-meaning prompt. Get your Legal and Privacy teams in the room before you pilot this one, not after.

Also worth flagging: this is where the conversation with Microsoft (Copilot), ServiceNow (NowAssist), and Salesforce (Agentforce) gets interesting. All of them want to be the “front door” for work. Your IT architecture review board is about to have a lot of meetings.

Flex Credits:The Math Nobody's Modeling

Time to do the math.

Keith Bitikofer put a slide up at the Philly RUG session with a figure that got the room’s attention: “$50-75 per run of the Agent??” Double question marks and everything. When I first saw it, I figured the number was a rumor, or an early estimate that hadn’t been fully vetted.

It’s not a rumor. It’s on the rate card.

The Workday Flex Credits Rate Card v262, last updated February 12, 2026, lists per-skill consumption rates for every agent. The public marketing slide that shows Self-Service at 1-5 credits and BP Optimize at 1 credit is accurate. It’s also a curated view. The full rate card includes rows that are meaningfully higher.

Here’s what got highlighted:

• Contract Negotiation Agent, full document review and redlining: 500 credits per document. At roughly $0.10 per credit, that’s $50 per contract.

• Recruiting Agent, Talent Rediscovery (Fetch): 750 credits per unique requisition. That’s $75 per req.

• Talent Mobility Agent, Internal Talent Visibility and Matching: 750 credits per unique requisition. Also $75 per req.

Compare that to Candidate Grading (Spotlight), which runs 6 credits per resume screen. Same Recruiting Agent, wildly different meters depending on which skill you invoke. The agent isn’t one thing. It’s a bundle of skills with very different price tags.

Run the math for a recruiting team at a mid-size company.

$45,000 a year, for one skill, on one agent. Fine.

Now layer in the platform meters.

API Requests: 0.006 credits each, or 60 credits per 10,000 calls. Integration Events: 0.25 credits each, or 25 credits per 100 events. Document Storage: 120 credits per GB stored annually, with a 10-credits-per-GB-per-month overage rate if you exceed.

Here’s the thing Keith made explicit that I hadn’t appreciated until I saw the deck: platform entitlements are “fair use” levels, not hard limits. When you exceed them, the overage draws down your Flex Credit pool. So your heavy integration month doesn’t just generate a separate line item. It eats into your agent budget.

But not every integration counts the same way.

Keith’s slide broke it down.

Doesn’t count toward entitlement (included with subscription):

• Workday Core Connectors

• Workday Everywhere (Slack/Teams)

• Integrations to and from Workday acquisitions (Hiredscore, Paradox)

• EIBs (for Integration Events only, not for API Requests)

• Workday Applications built by Partners

Counts (and overage consumes Flex Credits):

• Custom Studio integrations

• Orchestrations

• Document Delivery unless attached to an EIB

• Customer-built Extend apps

• External applications and AI Agents calling your tenant

• Certified Partner integrations (including EIBs) for API Requests

Translation: if you’re a heavy Studio shop, you’re exposed. If you have a sprawl of custom Orchestrations, you’re exposed. If you built out Extend apps, you’re exposed. Go look at what your integration footprint actually looks like before you assume anything.

A few safety valves worth knowing. Sandbox testing is free. Non-production consumption is shown in the console for informational purposes only, never billed. Traditional Workday “AI” and machine learning features (the non-agentic stuff that’s been in your tenant for years) don’t consume Flex Credits at all. The meter runs on agents and platform consumption, not on the AI features you already have.

If you do go negative on your entitlement, you don’t get cut off. Annual true-up at reconciliation. No service interruption.

“No service interruption” is not the same as “this will be cheap.”

Ask me how I know. When Justin was in second grade, he figured out how to switch the payment method on his iPad from password to his fingerprint, and proceeded to rack up a thousand dollars in Roblox in-game purchases. I found out about it a day later. Apple didn’t cut him off. The system worked exactly as designed. The damage was done before anyone noticed.

Flex Credits work the same way. If nobody is watching the meter, the meter doesn’t care.

Keith told us the Platform Consumption Console (PCC) is where you watch all of this. Credit balance, active entitlement, rate card view, drill-down on Sana and Agent usage. If you’re evaluating any of this, know where the PCC is before you turn anything on. Watch it like a hawk.

My honest read: the public narrative around Flex Credits has been “pay for what you use, usage-based, very reasonable.” The full rate card tells a more complicated story. For admins and HR self-service, the economics are fine. For talent and contract agents, per-run costs are substantial, especially at scale. For shops with heavy Studio integration footprints, the platform side is going to bite before the agent side does.

Get your CSM to run an estimate against your actual tenant volume before you commit to anything. Make sure your Finance counterpart sees the full rate card, not the marketing slide. And if you’re a recruiting-heavy shop considering Fetch or Internal Talent Mobility at scale, model the annual cost before you green-light the pilot.

What I’d Actually Pilot First

Not a ranking. A sequence.

Get on the UMSA if you aren’t already. Nothing else on this list works without it.

Get ASOR configured and governed. Before you turn on any agent, make sure you understand how to manage them. This is not glamorous work. It is the work.

Start using the Deployment Agent today. It’s in Customer Central, it’s free, it doesn’t require the UMSA, and it will save your team real time. There’s no reason to wait.

Pilot the Self-Service Agent. Highest ROI, clearest user value, fastest to demonstrate impact, and the credit consumption at 1-5 credits per action is manageable.

Pilot the Frontline Agent, if you have a frontline workforce. Disproportionate value for underserved users.

Invest in your skills data before you turn on Talent Mobility. The agent is only as good as the foundation underneath it, and at 750 credits per requisition, you want signal, not noise.

Watch Business Process Optimize carefully. This one could be a sleeper. Worth the evaluation even if you don’t act on it immediately. At 1 credit per BP event, the economics are friendly.

Everything else: evaluate based on your specific business. Don’t let a vendor pitch drive your roadmap.

What’s Notably Missing

A rundown is only useful if it also tells you what isn’t on the map. A few agents I’d expect to see that haven’t shipped yet, or at least haven’t been named:

A Learning Agent. Curious, given that Sana Learning was a major piece of the acquisition. Workday Learning and Sana Learn are being positioned as complementary, but I’d expect a dedicated agent for personalized learning paths, content generation, and coaching within the next 6 to 12 months.

A Performance or Feedback Agent. This one is tricky, and I’m not sure “performance agent” is even the right framing. Performance management in Workday has always been where good intentions go to get buried under rating scales and due-date reminders. Slapping an AI on top of that stuff doesn’t fix the process. The shift I think is actually coming is bigger: performance stops being an annual event and starts being continuous feedback consumption and aggregation. 360 feedback, peer recognition, awards, signals pulled from meeting notes and email threads, all of it synthesized into something coherent. That’s a fundamentally different product than “help me write a review.” Workday hasn’t named that agent yet. When they do, it’ll matter a lot more than a draft-my-review button.

A Comp Agent. Compensation planning is another area ripe for agent help: market data, internal equity checks, recommendation drafting. Conspicuously absent.

Read the roadmap accordingly.

The Bottom Line

The Workday agent lineup is going to keep changing. Names will shift. Capabilities will expand. A vendor you haven’t heard of yet will get acquired, and whatever they built will get absorbed into this roster within a year of the deal closing.

What won’t change is the architecture. Agents grounded in trusted data. Governance through ASOR. Security that inherits from your tenant. The pieces that actually determine whether this stuff works.

What also won’t change is the economic model. Flex Credits are here to stay. Know your rates. Know your meters. Know what counts toward your entitlement and what doesn’t. The agent roadmap is exciting. The consumption math is what’ll show up on your Finance team’s invoice.

Learn the architecture. The product names will take care of themselves.

The Illuminate branding lasted about eighteen months. Sana will probably last longer, but not forever. What you build on top of the real plumbing underneath: that’s the part that compounds.

-- Mike

P.S. A HUGE shout-out to Keith Bitikofer for the Philly RUG session that sharpened most of the Flex Credits analysis in this post. Keith pulled back the curtain on the parts of the rate card that weren’t in the marketing deck, walked the room through what actually counts toward entitlement, and demoed the PCC live. That’s the kind of practitioner-to-practitioner knowledge transfer that RUGs exist for. If your local RUG has a Keith, show up. If it doesn’t, please consider being one.

P.P.S. Ultimately, Roblox gave me my money back, but also perma-banned Justin’s original account. To this date, I still have not been forgiven. He apparently purchased the Titanic in some game where a random player becomes a Jaws-like great white shark and eats boats (and other players). I will never understand Gen-alpha. Then again, we had Doom. We’re definitely more screwed up.

tl;dr:

Moving from "AI that suggests" to "AI that acts" isn't an upgrade. It's a fundamentally different risk profile, and most HR tech teams have zero infrastructure for it.

Back in December, I published a Safety Checklist. Five questions to ask your vendors before buying. This is the harder list: three questions you have to answer yourself before deploying, plus the foundational data problem underneath all of it.

Your current change management process was designed for humans. Agentic AI is going to look at it, politely nod, and then set it on fire.

The Promotion Nobody Voted On

In Part 2, I introduced you to my digital intern. Permanently caffeinated. Never asks for PTO. Occasionally confidently wrong about whether the Egyptians invented pizza.

You liked the intern. I liked the intern. The intern was great.

Here's the problem: someone in your leadership team just watched a vendor demo where the intern wasn't fetching coffee anymore. The intern was approving purchase orders. The intern was adjusting compensation ranges. The intern was making decisions about people's careers while a sales engineer narrated the whole thing like a nature documentary.

And now that leader is in your office asking: "Why aren't we doing this?"

The intern didn't get smarter. The intern got promoted. And nobody in your organization voted on it, built a job description for it, or thought about what happens when the newly promoted intern does something catastrophically wrong at 2am on a Saturday with no human in the loop.

That's what this post is about.

The Jump

The distance between "AI that recommends" and "AI that acts" is not a step. It's a canyon.

On one side, you have a tool that says, "Hey, based on the data, you might want to consider adjusting this comp range." A human reads it, thinks about it, and decides. That's a suggestion. That's a really smart sticky note.

On the other side, you have an agent that adjusts the comp range. Autonomously. Based on logic that someone approved at some point, theoretically, probably during an implementation meeting where half the room was checking email and the other half was trying to figure out the conference room AV system.

Most HR tech teams have spent decades building change management processes, approval chains, and governance structures for humans making decisions. Humans who hesitate. Humans who call their colleague and say, "Does this look right to you?" Humans who have a gut feeling that something is off, even if they can't articulate why.

Agents don't hesitate. Agents don't have gut feelings. Agents have logic, and they execute that logic with the enthusiasm of a golden retriever who just discovered the gate is open.

Nobody has built a governance model for that. And the vendors selling you agentic capabilities are not going to build it for you. That's not a knock on them; it's not their job. It's yours.

So let's build it.

The Checklist Was the Entrance Exam. This Is the Course.

Back in December, I published a five-point AI Safety Checklist. These were the questions you ask your vendor before you buy anything. Data retention. Bias audits. Explainability. Security inheritance. Kill switches.

That was the entrance exam. Those questions protect you from buying the wrong thing.

This is different. These are the questions you have to answer about yourself before you deploy the right thing. Because you can pass every vendor checkpoint with flying colors and still crash the car if you haven't built the road.

Three questions. They sound simple. They are not.

1. "Who Gave It Permission? Does That Permission Still Mean What They Think It Means?" (Authority)

Every agent acts on authority. Someone, somewhere, at some point, said "yes, you can do this."

The question is: who? And when? And did they understand what they were saying yes to?

Because here's what happens in practice. During implementation, someone approves a capability called "Talent Optimization." It sounds great. Everybody nods. It goes into the statement of work as an enabled feature. Eighteen months later, the vendor has expanded what "Talent Optimization" means. Now the agent is auto-adjusting job posting language based on candidate conversion data. It's rewriting your job descriptions. In real time. Based on what gets clicks.

And nobody remembers approving that. Because technically, they didn't. They approved a category. The vendor filled in the details later. And the details now include an AI rewriting your EVP messaging to optimize for engagement metrics that may or may not align with what your talent acquisition strategy actually says.

This happens all the time with regular software. Feature creep is not new. But feature creep with a deterministic system means someone eventually notices and opens a ticket. Feature creep with an autonomous agent means the agent has been making decisions for weeks before anyone realizes the scope drifted.

The fix: A living authority register. Not a one-time implementation sign-off. A document (a real one, not a slide deck someone presented once at a steering committee meeting and then buried in a SharePoint folder called "Archive - DO NOT DELETE - Final v3") that says exactly what each agent is authorized to do, who approved it, when it was last reviewed, and what the boundaries are. Reviewed quarterly at minimum.

I know. Quarterly reviews. I can feel you closing the browser tab. Stay with me.

2. "Who's Watching What It Did? Can They Explain Why It Did It?" (Oversight + Explainability)

In the Safety Checklist, I told you to demand explainability from your vendors. A "Glass Box," not a black one. Every output should come with a reasoning chain.

That was the vendor's job. This is yours: who on your team is actually reading those reasoning chains?

Agents act fast. Humans review slow. If your agent runs daily and your review cadence is monthly, you have 30 days of unaudited decisions compounding on top of each other. That's not a gap. That's a canyon with a gift shop at the bottom selling "I Survived Our AI Governance Program" t-shirts.

The scenario: An AI agent is recommending compensation adjustments across open requisitions based on market data, pipeline conversion rates, and internal equity benchmarks. It runs every day. It touches 200 reqs over three weeks.

Recruiters see the recommendations. They look reasonable. They have a little confidence score next to them, which (as we discussed in Part 3) makes humans go absolutely limp with compliance. "The system says 87%? Must be right." They adjust the comp ranges. Hiring managers approve. Offers go out.

Three weeks later, someone in Total Rewards finally reviews the output. They discover the market data feed was pulling from a dataset that included contract roles, which inflated the benchmarks by 12%. Every comp adjustment for the last three weeks was built on a bad foundation.

You are no longer auditing a recommendation. You are unwinding a decision chain. Offers have been extended. Some have been accepted. And here's the part that should make your legal team's ears perk up: if those inflated benchmarks skewed differently across demographics (and they almost certainly did, because market data for contract roles doesn't distribute evenly across job families), you may have just created an adverse impact pattern. Not because anyone intended to discriminate. Because nobody was watching closely enough to catch the drift.

That's not an AI bias problem in the way the Checklist discussed it. You already asked the vendor for the bias audit. They passed. This is a monitoring problem. The model was fine. The data feeding it shifted. And nobody caught it because nobody was assigned to catch it at the speed the agent was moving.

The fix: The review cadence has to match the action cadence. If the agent acts daily, someone (a human, with judgment, and the authority to say "stop") reviews daily. And that person needs to be able to read the reasoning chain — not just see the output, but understand why the agent made the call it made. If your reviewer can't explain the agent's logic to a skeptical VP in plain English, the review is theater.

3. "What's the Damage Protocol?" (Rollback)

The Safety Checklist gave you the kill switch — the vendor-side toggle that lets you shut it off. Good. You need that.

But turning it off is step one. What about the damage that's already done?

Your agent is going to make a mistake. It is a statistical certainty. And I need you to internalize something: the mistake will not announce itself. There will be no pop-up that says "ERROR: I JUST DID SOMETHING DUMB." The agent will execute the wrong action with the exact same confidence it executes the right one. That's the whole thing about machines. They don't sweat. They don't pause. They don't get a weird feeling in their stomach when something doesn't add up.

You know who does get a weird feeling in their stomach? Brenda in Payroll. And Brenda is going to catch this. At 4:47pm on a Friday. And Brenda is going to send you an email with the subject line "Question???" and three question marks is never good. Three question marks means Brenda found a body.

The scenario: An agent flags a batch of employees for a compliance-related action based on certification expiration dates. The field hasn't been audited in eight months. Some of the dates are stale. Some are flat-out wrong because someone fat-fingered an entry during a mass upload and nobody caught it because (and this is the part where I stare directly into the camera) nobody audits EIB data after the upload.

The agent doesn't know the data is stale. The agent sees an expired certification, matches it to a business rule, and fires the action. Forty employees get compliance notices. Twenty of them have valid, current certifications. They are, understandably, upset. Some of them call HR. Some of them call their manager. One of them calls a lawyer. (There's always one. If you've worked in HR Tech long enough, you know that one person who has a lawyer on speed dial like it's a pizza place.)

And here's where the regulatory landscape makes this worse: depending on the nature of the compliance action and your jurisdiction, you may have just triggered a notification obligation. NYC Local Law 144 requires notice and disclosure when automated tools are used in employment decisions. The EU AI Act classifies HR systems as "high-risk." You didn't just send a wrong notice; you potentially sent a wrong notice using a process that regulators are actively watching.

The fix: A rollback playbook for every agentic capability you enable. Not "open a ticket." A documented protocol that answers: what gets reversed, who authorizes the reversal, how do we communicate to affected employees, who owns the post-mortem, and what's our regulatory notification obligation? That protocol needs to exist before the agent goes live. Treat it like a disaster recovery plan. Because that's what it is.

The Floor Under the Agent (Data Foundation)

I've saved the most boring and most important part for last.

Your agent inherits every sin in your data.

Every stale supervisory org. Every orphaned position. Every job profile that hasn't been audited since the original implementation when someone said, "We'll clean that up in Phase 2," and Phase 2 never came because Phase 2 is a myth. Phase 2 is the HR Tech equivalent of "we should get together sometime." It's never happening.

The agent doesn't know your data is bad. The agent treats your data as ground truth. It walks across your data foundation like it's solid concrete. If there are holes, the agent falls through them at machine speed. And it doesn't even know it's falling. It just keeps executing, confidently, all the way down.

Bad data in a traditional system creates bad reports. Bad data under an autonomous agent creates bad actions. Reports you can correct. Actions you have to undo. And some actions (an offer extended, a notice sent, a compliance flag triggered) are really, really hard to undo.

The minimum before you go agentic:

• Supervisory org audit. When was the last time you verified that your org structure in the system matches reality? If the answer involves the phrase "during implementation," that's not an answer. That's a confession.

• Job profile hygiene. Are your job profiles current, consistently structured, and actually reflective of the roles people are doing? Or are they a museum exhibit of what someone thought the role was in 2019?

• Data field freshness. Every field the agent will read needs a last-verified date. If you can't tell me when a data point was last validated, the agent shouldn't be acting on it.

• Integration integrity. If the agent is pulling data from multiple systems, are those systems in sync? Or are you running on "Schrödinger's Data" (a concept I covered earlier in this series) where the same employee exists in three systems with three different job titles and nobody knows which one is authoritative?

You don't need perfect data. (Perfect data is another myth, right up there with Phase 2 and "we'll migrate the historical data later.") You need audited data. You need to know where the holes are before the agent starts walking.

Your Change Management Process Is Kindling

I promised in Part 3 that we'd talk about why your current change management process is going to look at agentic AI and burst into flames.

Here's why.

Your change management process was designed for a world where a human is the decision-maker at every step. A change gets proposed. A stakeholder reviews it. An approval chain fires. Someone communicates it to the affected population. Training happens. Feedback gets collected. The cycle repeats.

That entire model assumes one thing: time. Time for review. Time for feedback. Time for someone to say, "Wait, I don't think this is right." Time for Brenda to get that feeling in her stomach.

Agentic AI compresses time to near-zero. The agent proposes, evaluates, and acts in the space between your first sip of coffee and your second. There is no "review window." There is no "let's run this by the team." The agent ran it by itself, and it approved unanimously (because it was the only one in the room, and it doesn't understand the concept of a dissenting opinion).

You don't need a new version of your change management process. You need to acknowledge that the old one was designed for a species that moves at human speed, and then build something new for a world where decisions happen at machine speed.

What does that look like? It looks like pre-approved decision boundaries instead of case-by-case approvals. It looks like automated exception detection instead of manual review queues. It looks like real-time monitoring dashboards instead of monthly steering committee decks. It looks like treating your governance model as a living system, not a binder on a shelf.

It looks like work. A lot of it. I'm sorry. I wish I could tell you there's a shortcut. The shortcut is called "let the agent do whatever it wants and hope for the best," and the technical term for that strategy is "career-limiting."

The AI hasn't changed. The models are getting better, sure, but the fundamental dynamic is the same as it was in Part 2 when I was using it to draft user instructions and untangle calc fields. It's a tool. A very fast, very confident, occasionally wrong tool.

What changed is the authority. You moved the intern from the desk next to yours (where you could see the screen, check the work, and catch the errors) to a corner office with signing authority and a door that closes.

The governance model is how you make sure that promotion is earned. Incrementally. With guardrails. With review cadences that match the speed of the decisions being made. With rollback plans that exist before you need them. With data foundations that have been audited this calendar year.

Not all at once on a Tuesday afternoon because someone in the C-suite saw a demo and sent you an email with "THOUGHTS??" in the subject line.

Build the governance. Audit the data. Match the review speed to the action speed. Plan the rollback before you need it.

And for the love of everything, answer the questions before you flip the switch.

That's the series. Four (okay, it ended up being five) parts. The vocabulary, the plumbing, the literacy gap, and the governance. If you've made it through all of them, you're better prepared than most of the HR tech teams I talk to. And if you read the Safety Checklist back in December, you've now got both sides: the questions you ask them and the questions you answer yourself.

If this series helped you think differently about AI in your environment, share it. Forward it. Post it. Do the LinkedIn screenshot thing. Build the governance your organization needs before someone else builds urgency you can't control.

And if you're sitting there thinking, "This is great, Mike, but how do I actually sell this to my leadership?" Yeah. I hear you. I might have another post in me on that. No promises. (Okay, soft promise.)

— Mike

Director HR Tech | Intern Supervisor

The Department of First Things First. For the people who do the work.

P.S. Justin asked me what "governance" means. I told him it's like the rules in his video game that keep other players from cheating. He said, "So... it doesn't work?" Kid might have a future in consulting.

You just found The Department of First Things First. Maybe someone forwarded you a post. Maybe you caught me on Unsafe Harbor or Workday Gold. Maybe you Googled “why does Workday hate me” and ended up here.

Either way, welcome. Here’s the quick version of what this place is:

I’m Mike Domingo. I’ve been in HR technology for over twenty years. I’m currently a Sr. Manager of HR Technology at CVS Health. I co-organize the Philadelphia Workday RUG. I have a PMP, a DASM, and four Workday Pro certifications. I have opinions, and this is where I put them.

This newsletter is for the people who actually configure, architect, and maintain enterprise HR systems. Not the analysts. Not the buyers. Not the people who attend conferences and nod. The people who go back to their desks on Monday and have to make it work.

I publish weekly. No paywalls. No sponsors. No vendor cheerleading.

Here are five posts that will tell you whether this is your kind of place:

1.

My son Justin asked me for $10 to buy a cosmetic skin for his video game character. It didn’t make him faster or stronger. He just looked cool while falling off a cliff. Then Workday released their Sana-driven UI and I realized they’d done the exact same thing for a billion dollars. This post breaks down what the new UI actually means for practitioners, where the security nightmares hide, and why a search bar on top of a messy warehouse doesn’t clean the warehouse.

2.

Josh Bersin published a piece about AI “orchestration” that made it sound like the future was already here. I wrote the practitioner’s rebuttal. This one hit a nerve — 21,000 LinkedIn impressions, 233 reactions, 44 saves, and it doubled the newsletter’s subscriber count in a week. If you’ve ever sat in a meeting where an executive waved an analyst article at you and asked “why aren’t we doing this yet?” — this post is the answer you wished you had.

3.

Bersin argued that enterprise software is protected by a moat of decades-deep business rules. He’s right about the moat. But for the people actually using the system, that moat feels like a ten-lane tollbooth. This post introduced the framework I keep coming back to: the ERP isn’t going away. But the future is the ERP becoming headless infrastructure, buried so deep under experience layers and AI agents that the end-user forgets it exists.

4.

The hardest part of moving from Workday analyst to architect isn’t learning a new module. It’s the psychological shift from fixing tickets to navigating politics. I break down what that transition actually feels like: the muscle memory trap, the instant gratification desert, and why the Kitchen Table Test (can you explain your architecture to an 11-year-old?) is the only metric that matters.

5.

I Let My 11-Year-Old Audit Workday’s UI. He Lasted 3 Minutes.

I handed my son Justin a Workday tenant and asked him to request a day off. He’s a digital native who can navigate any app in seconds. He lasted three minutes before giving up. If a kid who grew up with an iPad needs a Quick Reference Guide to use your system, your design is broken. This post is the one that makes non-technical leaders finally understand the UX problem.

If any of those landed, you’re in the right place. Subscribe and I’ll show up in your inbox weekly with architecture takes, vendor accountability, and the occasional story about my son accidentally explaining enterprise strategy better than a $50,000 consultant.

— Mike

Director HR Tech | Keeper of the Layer Cake

P.S. If you want to hear me talk instead of read me type, I’ve been a guest on:

• Talent Experience Live with Devin Foster at Phenom — “Speed Meets Experience” (to be published)

• Unsafe Harbor with Christian Delcid — Episode 14

• Workday Gold with Keith Bitikofer — December 19th, 2025

All are worth your commute.

tl;dr: Josh Bersin released a podcast about AI literacy last week. It’s fourteen minutes of anxiety creation and product placement. His core argument — “AI learns from language, so be logical” — isn’t wrong, it’s just useless for practitioners. Real AI literacy for HR tech people isn’t about prompting or vocabulary. It’s about governance: who holds the leash on AI agents, what happens when they act on employee data, and how the requirements scale dramatically across Workday Illuminate’s tiers (Accelerate → Assist → Transform).

Josh Bersin dropped a podcast last week called “What Does AI Literacy Really Mean?”

Fourteen minutes long. And in those fourteen minutes, he managed to identify a problem, create anxiety about it, and sell you the cure. It’s a clean structure. The kind you’d respect if it weren’t aimed directly at your budget.

I’m not saying AI literacy isn’t important. It is. But when an analyst defines the disease and sells the medicine, you’re allowed to ask for a second opinion.

Here’s mine.

“AI Learns From Language” Is Not a Framework

Bersin’s big reveal is that AI learns from language, so you need to be able to put your needs into “logical statements” to get the most out of it.

That’s like telling someone learning to drive that cars run on gasoline. True. Not particularly useful when you’re trying to parallel park.

You know what AI literacy actually looks like for the people reading this newsletter? It looks like understanding why the same prompt gets wildly different results depending on how much context you give it. It looks like knowing what RAG is — not because you need to build one, but because your vendor is about to sit across from you and say their product “uses RAG” with the confidence of a sophomore who Wikipedia’d the topic ten minutes before the presentation. You need to know whether that’s meaningful or marketing.

It looks like understanding the difference between a model suggesting something and a model doing something. Which (and I really need everyone to internalize this) is the difference between “helpful tool” and “thing that just changed an employee’s compensation while you were in a meeting about change management.”

That’s not “be more articulate.” That’s architecture.

Bersin doesn’t go there. And to be fair, that’s not really his audience. His audience is buying. This newsletter’s audience is building. The advice that works at 30,000 feet can be actively misleading at ground level.

The Grocery List Is Not a Meal

In his podcast notes, Bersin name-drops bias, risk, auditability, data quality, and tuning. He even throws in a note about “vibe coding” being more complex than people think.

Cool. That’s a grocery list.

You don’t hand someone a list that says “chicken, rice, lemon, garlic” and call yourself a cooking instructor.

Who owns bias review when an AI suggests a compensation adjustment? What’s your escalation path when an agent takes an action on employee data that doesn’t match your business rules? What do you do on the Monday morning when someone in legal calls because an AI-driven recommendation created a disparate impact pattern that nobody caught because nobody was assigned to catch it?

These aren’t abstract questions. They’re Tuesday. They’re the actual work. And no podcast or newsletter (mine included) answers them for you. The difference is I’ll tell you that upfront instead of pointing you toward a subscription product.

So What Does AI Literacy Actually Look Like?

It looks like governance.

I know. I’m sorry. I just said the word “governance” and half of you felt your soul leave your body. Stay with me.

Not the “let’s draft an AI ethics statement and put it on the intranet where it can live next to the recycling policy and the 2019 holiday party photos” kind.

The kind where you decide (before you turn anything on) who holds the leash, how long it is, and what happens when the dog runs into traffic. Because the dog will run into traffic. The dog always sees a squirrel named “efficiency” and just goes.

I wrote about this in “The Agentic Leash” a few weeks ago, and the response told me something: practitioners are hungry for this conversation. Not the theoretical version. The operational one.

So let’s get operational.

The Tiers Are Not Created Equal

If you’re in the Workday ecosystem (and if you’re reading this, there’s a solid chance you are), you already know about Illuminate. What you might not have thought about is how dramatically the governance requirements change as you move through the tiers.

Accelerate is the easy one. Embedded AI. Summarization, natural language search, smart suggestions that surface information. Low risk. The AI is basically a really enthusiastic research assistant who never sleeps and never asks for PTO. You still make every decision. Governance here is mostly “did we turn it on?” and “does our security model still make sense?” This is the tier where everyone feels smart. Enjoy it.

Assist is where it gets interesting. The AI starts making recommendations that humans act on. Job matching. Compensation suggestions. Talent pipeline insights. The AI is no longer just surfacing information — it’s interpreting it. And the human on the other end might not know how to evaluate whether the interpretation is any good.

That’s where literacy matters. Not “can you prompt well?” but “do you understand what you’re looking at well enough to push back on the machine?” Because here’s the thing about humans: we are phenomenal at deferring to something that sounds confident. The AI gives you a recommendation with a little percentage next to it and suddenly everyone in the room is nodding like it came down from a mountain on a stone tablet.

Transform is the one that keeps me up at night. This is where agents take action. Not suggest action. Take action. And the governance model for that is fundamentally different from anything most HR technology teams have ever had to build.

Who approved the model’s logic? Who reviews the output before it hits a manager’s inbox? What if it’s wrong? What if it’s right but the business context changed and nobody told the model? What’s the rollback plan?

These questions don’t show up in a 14-minute podcast about AI literacy. But they’re going to show up in your implementation, whether you’re ready for them or not.

A Scenario, Because I Know How We Learn

Let’s say your Workday environment is running an AI agent (Transform tier) that analyzes talent pipeline data and suggests changes to open job requisitions. Adjusted job descriptions. Revised compensation ranges. Maybe even suggested changes to hiring criteria based on what’s working in similar roles.

Sounds amazing. Sounds like the future.

Now: who approved the logic the model used to decide what “working” means? Is it optimizing for time-to-fill? Quality of hire? Retention at 12 months? All three? Because those optimization targets can conflict with each other, and the model picked one. Or blended them. And someone in your organization needs to know which, and why, and whether that aligns with what your CHRO actually wants (or what they said they wanted six months ago before the reorg).

The output lands in a recruiter’s queue. It says “adjust the comp range for this req from $85K–$95K to $92K–$105K based on market data and pipeline conversion rates.” The recruiter sees an AI-generated recommendation and thinks, “Cool, the system says so.” They adjust it. The hiring manager approves it. Nobody interrogated it.

Nobody asked: what market data? What pipeline? Over what time period?

Six months later, you’ve got a comp equity issue that nobody can trace because it started with an AI recommendation that twenty people approved by not questioning it.

That’s not an AI failure. That’s a literacy failure. And it’s a governance failure. Because nobody defined who’s supposed to ask those questions before the recommendation goes live.

This is what AI literacy looks like in practice. It’s not prompting. It’s not “being logical.” It’s understanding the system well enough to know when to trust it, when to question it, and when to shut it off and figure out what went wrong.

The Encyclopedia vs. The Architecture

Bersin will keep writing the encyclopedias. There’s value in that. The categories, the frameworks, the market maps…they serve an audience. He’s earned that lane.

But if you’re the person who has to actually build the thing? Configure the tenant? Explain to your VP why the AI did what it did? Stand in front of a governance committee and say “here’s how we control this” while someone from legal stares at you like you just described a haunted house?

You don’t need an encyclopedia. You need architecture.

AI literacy for practitioners isn’t a course you take or a product you subscribe to. It’s a muscle you build. It’s knowing your platform, knowing your data, and knowing who’s accountable when the machine makes a move.

Next week: what happens when you actually let the agent off the leash. Part 4 goes deeper on the governance operating model: who owns what, how to build the review cadence, and why your current change management process is going to look at agentic AI and simply burst into flames.

If this is the kind of thing that makes you nod at your screen in a way that concerns your coworkers, please subscribe. And if you know someone who’s building, not just buying, share this with them. Forward the email. Post the link. Do that thing where you screenshot a paragraph and put it on LinkedIn with a “THIS 👆” caption. I won’t judge.

(I will judge a little.)

— Mike

The Department of First Things First. For the people who do the work.

P.S. Justin just wrapped up travel 14U volleyball season, where the gym acoustics are specifically engineered to ensure no parent retains their hearing by age fifty. And now we’re rolling straight into spring tennis tournament season, where I get to watch twelve-year-olds make line calls that would get a FIFA referee fired. USTA rules: on the line is out, inside the line is questionable, and clearly in is “I didn’t see it.” Governance starts at home, people.

"Built in" is doing a lot of work. Workday's security model isn't two words. It's domain policies, BP policies, constrained groups, intersections — and none of that translates cleanly when an agent pulls Slack data into a Workday comp profile.

"Thousands of agents" is a loaded gun, not a feature. Who approves them? Who audits them? What happens to API performance when 1,000 custom agents start polling Workday at once? Josh doesn't ask. You have to.

ASOR gets one mention. In parentheses. For a guy who coined "Intelligent Orchestration," the actual orchestration layer gets less screen time than the Galileo sales pitch.

The Copilot collision Josh skipped. Your CIO already paid for Microsoft Copilot. The question isn't "is Sana better than Copilot" — it's "is Sana better enough to justify funding two AI front-ends after we told the board we were consolidating."

Disclosure ≠ objectivity. Telling me you own the stock doesn't make the analysis objective. It just tells me why it isn't.

The Form Check. Six things your team should be interrogating right now — security inheritance, agent approval pipelines, data lineage, hallucination monitoring, migration paths, and real TCO (not sticker price).

It's 2:00 AM. You can't sleep. You're flipping channels.

And there he is. Perfectly lit. Radiating confidence. Leaning into the camera with the energy of a man who has personally tested this product and needs you to know about it.

The demo is flawless. The stains disappear. The before-and-after is stunning. Four amazing product features, each one bigger than the last. And right on cue — the call to action:

"Call us if you want to walk through the details."

I had a Billy Mays moment this week. You know the one. Where you're watching what looks like industry analysis, and then the camera pulls back and you realize…Oh! This is an infomercial. And the host is really good at it.

The Disclosure That Changes Everything (And Nothing)

To his credit, Josh tells you the deal right up front in paragraph one. He's a Sana customer. He's a Sana partner. His product, Galileo, runs on Sana. He's been using it for three years.

That's honest. I respect that.

But here's the thing about disclosure: it's necessary, but it's not sufficient. Telling me you own stock in the company doesn't make your stock analysis objective. It just means I know why it isn't. And once you know that Josh Bersin is reviewing his own landlord's renovation while selling condos in the building, every sentence reads differently.

Let's walk through his four big announcements the way an enterprise practitioner hears them. Which is: differently.

Announcement 1: All Workday customers get access to Sana for Workday.

Sounds incredible. What does "access" mean?

Is it bundled with existing SKUs? Is it a freemium play with an upsell? Is there a per-user cost? Per-transaction? Does my current contract cover it or is this a massive change order?

Josh doesn't say. A CHRO doesn't ask. But you — the person who manages the vendor relationship, reads the SOW, and explains the budget impact to the CFO — you're already reaching for your Workday rep's number.

(They're not going to pick up. The pricing is still "TBD." You know this.)

Announcement 2: Sana Enterprise connects to Salesforce, Teams, Slack, and more.

Multi-system integration through a single AI front-end is genuinely interesting. But here is what Josh calls Workday's security model: "built in."

Two words. For the most complex permission framework in enterprise HR technology. Domain security policies. Business process security policies. Constrained and unconstrained security groups. Intersection configurations.

"Built in." That's like saying OxiClean works on "all fabrics." Technically, the box says that. But you and I both know there's a reason the demo always uses a white t-shirt and never your grandmother's silk blouse.

But the real nightmare isn't just Workday's security; it's security translation. When Sana pulls a Slack message and pairs it with a Workday comp profile, whose security model wins? If an agent uses an Integration System User (ISU) to connect Workday and Salesforce, how does it respect field-level security across two entirely different data models?

The infomercial shows someone asking Sana for their PTO balance and getting a clean, happy answer. It does not show the data breach review after an agent surfaces comp data to someone in a view-only job requisition role because the integration couldn't translate the constrained security group.

Announcement 3: You can build your own agents in Sana.

Josh says: "Companies will build thousands of these agents."

Thousands.

My son, His Highness Justin (almost 12), can spin up a custom Roblox server with custom mods in ten minutes with zero oversight. Do I want an enthusiastic compensation analyst building a Workday agent with that same level of friction?

Who monitors for hallucination? Who audits data access patterns? What's the version control story? What happens when that comp analyst inadvertently exposes salary band data because the security inheritance isn't what they assumed? Or worse, what happens to your system performance when 1,000 custom agents start continuously pinging Workday's APIs to check for status updates?

Josh describes a loaded gun and spends the whole segment talking about how shiny it is.

The safety doesn't get mentioned. Not once.

Announcement 4: Sana's AI infrastructure becomes Workday's AI infrastructure.

Josh says the Illuminate brand is essentially being absorbed. New agents will be built within Sana's infrastructure going forward. Cool.

But what happens to the governance framework? Workday just spent a year training the ecosystem on how to govern Illuminate via the Agent System of Record (ASOR). Now they bought a completely new engine. Does the ASOR governance framework survive the transplant, or are we back to square one?

Want to know how seriously Josh takes ASOR? He mentions it once. In parentheses. Literally in punctuation designed for afterthoughts. He calls it "critical"…and then treats it like a footnote. For a guy who coined "Intelligent Orchestration," the actual orchestration layer gets less screen time than the Galileo sales pitch.

And when it comes to Paradox, HiredScore, and the rest of Workday's recent acquisitions? Josh's answer is: "will come out over time." If you're a customer who just finished a HiredScore integration, that sentence should make your eye twitch. Workday spent serious money acquiring these tools. You spent serious effort implementing them. "Will come out over time" is the analyst equivalent of thoughts and prayers.

The Boardroom Collision Josh Ignored

Josh's competitive landscape section is polite. Oracle gets two sentences. SAP/Joule gets a nod.

But there is a massive, existential boardroom collision coming, and Josh completely sidesteps it.

If you are an enterprise shop, you probably already have Microsoft Copilot. Not because you chose it, but because it came with the E5 agreement you signed two years ago. It's already in the budget. Already past procurement. Already blessed by InfoSec. Already in the SSO flow.

Now imagine walking into your CIO's office and saying: "I need budget for a second AI platform."

The CHRO wants Sana because it's "HR native." The CIO wants Copilot because it's already deployed and paid for. The question isn't "Is Sana better than Copilot?" Features comparisons are easy. You can build a slide for that.

The question is: "Is Sana better enough to justify paying for both?"

That is a "why are we funding two AI front-ends when we told the board we were consolidating vendor spend" conversation. And it's a conversation you can't objectively write when your own product runs on one of the platforms being evaluated. A practitioner actually has to live in that overlap. You have to answer the question your CIO will absolutely ask: "Can't we just build a Workday connector in Copilot Studio?"

Maybe the answer is yes, Sana is worth it. Maybe the Workday-native security layer and the learning platform integration genuinely differentiate it. But that argument has to be made with math, not with a Billy Mays demo. And the person who has to make that argument is you. Not Josh.

The Form Check: What You Should Actually Be Asking

Maybe Sana clears every one of these questions. Great. Let's see the test results. The point isn't that the answers are bad. The point is that nobody is asking the questions out loud, and somebody has to.

If Workday is serious about Sana as the agent infrastructure, here is what your team needs to be interrogating right now:

1. Security Inheritance Validation: Don't trust "built in" until you've tested it. Build test cases that intentionally cross security boundaries. Verify that Sana respects domain security, BP security, and field-level security independently across integrated systems.

2. Agent Inventory and Approval: If "thousands" of agents are coming, you need a pipeline. Not a free-for-all. Who approves them? Is there a review board? A sandbox environment? A staging process?

3. Data Lineage and Auditability: When an agent surfaces an answer, can you trace where that data came from? Which Workday report or data source did it pull? Was it real-time or cached? If an agent gives a wrong answer that leads to a business decision, can you reconstruct the logic?

4. Hallucination Monitoring: LLMs hallucinate. Period. What's the monitoring framework for agent-generated responses? Is there a confidence threshold below which the agent declines to answer? Who reviews flagged responses? Nobody has a great answer for this yet. That's exactly why you need to be asking.

5. Migration and Coexistence: If you've invested in Illuminate, what's the migration path? If you've built custom apps on Workday Extend, what's their future in a Sana-native architecture? Get it in writing. "Will come out over time" is not a migration plan.

6. Cost Modeling: What's the Total Cost of Ownership? Not just the sticker price, but the incremental FTE for agent governance, the change management burden, and the opportunity cost of not spending that budget elsewhere. Especially if you're already paying for Copilot.

The Laundry Room

Here's the thing about Billy Mays. He wasn't a con man. OxiClean actually works.

But Billy Mays didn't invent OxiClean. He was a paid endorser who did demos under perfect studio conditions. The lighting was right. The stain was pre-selected. And the camera cut away before you saw what happened when you tried it on the wrong material.

That's this article. Sana might be legitimately great. Josh clearly loves using it.

But this piece isn't strategic analysis. It's a guy in a blue shirt doing a demo with a white t-shirt and a perfect stain, under studio lighting, with a 1-800 number at the end.

The part the infomercial never covers is what happens when you get home, open the box, and try it on real fabric with real stains in your actual laundry room. That part has a name. It's called enterprise architecture.

And the stains are real.

-Mike

The infomercial ends. The laundry doesn't. Subscribe to The Department of First Things First.

This is a conversation space exclusively for subscribers—kind of like a group chat or live hangout. I’ll post questions and updates that come my way, and you can jump into the discussion.

Join chat

How to get started

1. Get the Substack app by clicking this link or the button below. New chat threads won’t be sent sent via email, so turn on push notifications so you don’t miss conversation as it happens. You can also access chat on the web.

Get app

2. Open the app and tap the Chat icon. It looks like two bubbles in the bottom bar, and you’ll see a row for my chat inside.

3. That’s it! Jump into my thread to say hi, and if you have any issues, check out Substack’s FAQ.

Last week, I gave you the vocabulary. RAG. MCP. Agents. The decoder ring for every vendor pitch and analyst article your CHRO will forward you at the moment.

But I also made you a promise: I'd show you what this actually looks like on the ground.

I spend a lot of time preaching about agentic governance and the risks of letting AI loose in your HRIS. But I also use these tools every single day. They're not optional anymore. They're how I survive.

So here it is. The truth about how I actually use AI at work. No vendor hype. No magic wands. Just the plumbing.

tl;dr:

I use AI every day, but I don't let it touch my Workday tenant. It's a digital intern — fast, tireless, occasionally confidently wrong. Three use cases that save me ~15 hours a week:

• The Translator: Drop dense Workday docs into an LLM with a specific prompt, get clean user instructions back in 15 seconds instead of 45 minutes.

• The Synthesizer: Feed scrubbed UAT feedback into AI, get the actual root causes instead of spending two days manually tagging a spreadsheet.

• The Wrench: Use AI as a pair-programmer for calc fields and XSLT — but make it explain its work so you can maintain it later.

• The rule: You are the Architect. The AI is the intern. Never forward the intern's first draft to the CHRO. Never paste its code into production without sandbox testing.

The Confession

Let's get this out of the way.

Right now, I do not let an autonomous AI push data into my Workday production tenant. The risk of corrupting foundational data is too high. Full stop. (That’s not to say this is a forever thing. But until I see things like ASOR mature, that’s my story and I’m sticking to it.)

So the AI doesn't run the system. It runs me.

I treat large language models (Claude, ChatGPT, whatever you prefer) as permanently caffeinated digital interns. They don't have the keys to the building. But they sit right next to my desk, and they help me translate, synthesize, and build the architecture that I eventually deploy.

Here are the three ways that intern earns its keep.

The Translator: Drafting User Instructions

If you work in HR Tech, you know that Workday release notes read like stereo instructions written by someone who has never plugged in a stereo.

A store manager doesn't have the patience for a 12-page PDF on absence management business process changes. They need to know what buttons to click so they can get back to their actual job.

The workflow: I take the dense Workday documentation (community, config docs, etc.) and drop it straight into my LLM with a specific prompt:

You are an expert technical writer. Read this Workday configuration document. Translate it into a simple, 3-step instructional flow for a non-technical manager. Use an encouraging, conversational tone. Strip out all system jargon — do not use words like 'Supervisory Org' or 'Condition Rule.' Output the result with bold headers and short bullet points.

What used to take me 45 minutes of staring at a blank screen now takes 15 seconds. The AI does the heavy lifting of translation. I tweak the final copy.

That's it. No magic. Just a really good first draft that I'd never have the energy to write from scratch at 4pm on a Thursday.

The Synthesizer: Making Sense of the Noise

After a major module rollout, the noise is deafening. Hundreds of UAT tickets. Angry emails. Help desk complaints that range from legitimate system bugs to "I clicked the wrong button and now I'm panicking."

When you're drowning in a sea of "This screen is broken!" and "I can't see my team!" tickets, stepping back to see the actual pattern is almost impossible.

The workflow: I scrub the feedback data of any PII, export the raw complaint logs to a CSV, and feed it to the AI (in this case, my work Copilot. Don’t use commercial tools for work stuff).

Analyze these 200 user feedback tickets from our recent performance rollout. Categorize them by core issue. Identify the top three root causes of user friction.

Within seconds, the AI cuts through the emotional noise. It tells me: "70% of these tickets aren't system errors — they're managers who don't understand the new security role mapping."

Instantly. A chaotic pile of complaints becomes an actionable punch-list.

That's not replacing my job. That's giving me X-ray vision into a dataset I'd have spent two days manually tagging in a spreadsheet.

The Wrench: Calc Fields and Code I Can't Write Alone

I have expert-level knowledge of Workday architecture. But I'm the first to admit my pure programming chops are light. Sometimes you hit a wall. You need to nest five calculated fields, or untangle a messy XSLT transformation on an inbound integration, and the syntax just isn't clicking.

This is where the AI becomes the ultimate pair-programmer.

The workflow: I don't just ask the AI to write the code. I ask it to teach me.

I am trying to build a nested calculated field in Workday to extract the length of service for employees in a specific supervisory org, but my logic is failing. Provide the correct structure, and explain exactly what you did so I understand the mechanics.

It's having a senior developer sitting over your shoulder who never gets annoyed when you ask "but wait, why?" for the fourth time.

It untangles the spaghetti. Writes the fix. And crucially — helps me understand what it did so I can maintain it later. Because nothing is worse than deploying a calc field you can't explain to the next person who inherits your tenant.

Worth knowing: The workflows above use general-purpose LLMs (Claude, ChatGPT, whatever you prefer.) But if your team needs something purpose-built for Workday, check out Mando. It's a Workday-specific RAG tool with curated content from vendors and ecosystem experts, every answer cites its source, and it gets smarter to your org over time. Great for config questions, troubleshooting, and getting new analysts up to speed fast. (They also published a solid AI 101 piece on the Customer Sharing Movement blog — go read it.) (NOTE: This is a personal advocacy only. I have no arrangement with Mando, I just appreciate their work)

You Are the Architect. The AI Is the Intern.

His Highness, Justin - who is nearly 12 now and requires the formal title - thinks my job consists entirely of "typing."

He isn't entirely wrong.

A massive part of HR tech leadership is translating complex technical realities into language that non-technical people can act on. My digital intern saves me roughly 15 hours a week on brainstorming, drafting, and untangling logic. It lets me punch way above my weight class.

But here's the rule, and it's unbreakable:

You never take the intern's first draft and forward it directly to the CHRO.

You never paste the intern's calculated field into the production tenant without testing it in sandbox first.

Use the AI to build the engine. Let it write the drafts, synthesize the feedback, map the logic. But you are the one who turns the wrench, pays the integration tax, and owns the final result.

The intern is fast. The intern is tireless. The intern is occasionally confidently wrong about whether the Egyptians invented pizza.

That's why you're still in the chair.

Next up: we're going to talk about where this gets scary…what happens when the intern gets the keys to the building. Agentic AI governance, the risks nobody's talking about, and why your foundational data strategy is the only thing standing between "helpful assistant" and "unsupervised chaos." Stay tuned.

And if you missed Part 1 (AI for HR People Who Don't Want to Sound Dumb in Meetings) - start there. Consider it the prerequisite.

If this was useful, share it with someone who's still writing user instructions from scratch at 4pm on a Thursday. They need this.

— Mike

Every Director of HR Technology knows the feeling.

Your inbox chimes. It's an email from your CHRO containing a link to the latest visionary article from a top-tier industry analyst, usually pitching some new "Expert HR AI."

The subject line is always the same: “Have you seen this? Why aren’t we doing this yet?”

I was explaining one of these massive AI platform announcements to my 11-year-old son, Justin, the other day (whilst getting my butt kicked at ping-pong). I told him they had loaded 25 years of HR research into an AI that could answer any policy question. He thought about it for five seconds and said:

"So it's basically the Wikipedia of work?"

Out of the mouths of babes. That is exactly what it is.

And it's why these articles make my blood boil. They sell the dream of frictionless AI but completely ignore the architectural plumbing required to make it actually execute tasks in Workday.

So let's fix that. Every major AI concept that matters for HR right now can be explained through the lens of an 11-year-old doing homework on a Sunday night (AKA - the Night of Tears and Regret). And if it can't survive that test, it's not worth your time.

Let’s get down to it.

The Large Language Model (The Kid)

Before we get into tools, let's start with the thing itself.

A large language model (LLM), like ChatGPT, Claude, or Gemini, is like Justin after 11 years of school, YouTube, and life. He's absorbed a ton of information. He can reason through problems. He can even surprise you with what he knows.

But he's working from memory. Sometimes that memory is sharp. Sometimes it's... creative. ("Dad, I'm pretty sure the Egyptians invented pizza.")

The model is the brain. Everything else we're about to talk about is how you make that brain useful in your world.

Prompt Engineering (The Homework Instructions)

Here's where most people start, and honestly, it's the highest-ROI concept on this list.

You know how if you tell Justin "do your project," you get a half-hearted paragraph and a shrug? But if you say "write two paragraphs about daily life in ancient Egypt, one about food and one about housing, with at least one specific detail in each" - Suddenly you get something you can put on the fridge?

That's prompt engineering. It's just being specific.

Tell an AI "write a job description," you get LinkedIn boilerplate. Tell it "Write a job description for a Workday Report Writer, emphasizing cross-functional collaboration and calculated fields, in a tone that's professional but not robotic" - Now we're cooking.

Most people blame the AI when they get bad output. Usually, it's the prompt.

RAG: Retrieval-Augmented Generation (The Study Guide)

Justin needs to write about Egypt, but his memory is fuzzy. So you go full research mode, highlight the good parts of a few articles, drop the stack on his desk, and say "here, work from this."

Now Justin can write something solid. But here's the catch: he can only work with what you gave him. If his teacher posted updated instructions to Google Classroom an hour ago, he doesn't know. He's working from a frozen-in-time stack of paper.

That's RAG. In the real world, RAG means you feed your company's actual documents (policies, SOPs, knowledge base articles) to the AI at query time. It doesn't hallucinate an answer. It pulls from your stuff.

This is what Justin called "the Wikipedia of work."

And when an analyst sells you a massive "Expert AI" for HR? (Ahem, Galileo) That's what they're selling you. A RAG model. It's study guide.

But it's read-only. It knows the answer. It can't press the buttons.

MCP: Model Context Protocol (The Badge and the Keys)

Same Sunday night. Same project. But instead of printing out articles, you take a different approach.

"Here's the login to Google Classroom. Here's the school library database. And here's your teacher's email if you need to ask a question."

Now Justin doesn't need you to pre-chew everything. He can see the latest assignment instructions himself. Check if the rubric changed. Search the library. Even email Mr. Hottenstein to clarify a question (shout out to Mr. Hottenstein, the GOAT). He's got live access to actual systems, and he can do things in them. Not just read stuff you handed him.

That's MCP.

MCP stands for Model Context Protocol. If you want another analogy, think of it as USB-C for AI. Remember when every phone had a different charger? MCP is the universal adapter; an open standard that gives AI a consistent way to plug into your tools. Your calendar. Your HRIS. Your ticketing system. Without someone building a custom integration for each one.

Where RAG gives AI a cheat sheet, MCP gives AI an employee badge.

RAG is read-only and point-in-time. MCP is live, and it goes both directions; the AI can read and write.

AI Agents (The Self-Directed Kid)

Now let's level Justin up.

Instead of you standing over his shoulder going "OK, now open the textbook... now write a topic sentence... now check the rubric..." Imagine Justin just handles it. He reads the assignment, makes a plan, looks up what he needs, writes a draft, checks it against the rubric, and revises. All on his own (AKA - Dad Nirvana). He might ask you a question along the way, but he's driving.

That's an AI agent. An LLM that can plan and execute multi-step tasks: not just answer a question, but figure out what steps are needed, use the tools available, and work through the problem.

In HR terms: asking "what's our PTO policy?" is RAG. Asking "look up my PTO balance in Workday" is MCP. But an agent handles the whole thing: "I need two weeks off in August. Check my Workday balance, find dates that don't conflict with my team's schedules, and draft the request for my manager's approval."

Plan. Research. Execute.

This is the direction everything is heading. And it's why MCP matters so much. Agents need secure, structured access to your systems to do anything useful. MCP is the infrastructure that makes that possible.

We're early. But not as early as you would think. This is live and in the wild, not yet everywhere, but we're seeing it. The groundwork you lay now (clean data, solid integrations, well-defined business processes) is what determines whether your org is ready when agents show up for real.

The Cheat Sheet

Because I know you skimmed. (It's fine. I did too.)

(Format: Concept | Justin Analogy → HR Example)

• LLM | The kid (smart but working from memory) → The AI itself: Claude, ChatGPT, etc.

• Prompting | Clear homework instructions → Writing better queries to get better output

• RAG | The highlighted study guide → AI that answers from your actual policy docs

• MCP | Login access to school systems → AI that connects to Workday/ServiceNow to take action

• Agents | Doing the project end-to-end → AI that handles multi-step workflows autonomously

How to Answer Your CHRO

You don't need a CS degree to make good decisions about AI in HR. But you do need this vocabulary, because the vendors are coming, the RFPs are coming, and you do not want to get swindled by someone selling a RAG model disguised as an agent.

The next time an executive forwards you an analyst's hype piece and asks "Why aren't we doing this?", you have your answer:

"The tool in this article is a RAG model, a great study guide, but it doesn't execute business processes in Workday. Our primary mandate right now is fixing our foundational data so we can eventually deploy true agentic AI that actually removes friction for our managers. Buying another study guide won't fix the workflow."

Let the analysts write the encyclopedias. We have architecture to build.

This is the first piece in a series on how the Department of First Things First actually uses AI in its workflow. Not theory! The real, messy, practical version. If you want to follow along, subscribe and I'll see you next week.

PS- If this was useful, share it with someone in your org who keeps nodding along in AI conversations but definitely Googles everything afterward. We've all been there.

PPS- I have just received a directive from Justin's agent (his mother) that in order for Justin to maintain his position as Chief UX Tester at the Department of First Things First, I am hereby required to refer to him as "the nearly 12-year-old." I have acquiesced to this demand without negotiation. Some agents don't need MCP to execute.

— Mike