Read time: 6 minutes

A collaboration with Geoff Momin from Interlay.

What a decade this year has been!!! 😉

Since the Rising 2025, Workday has made a lot of AI-focused moves. Acquisitions, renames, a whole new pricing model, varying levels of hype, and varying levels of success.

Most of us can name at least a few of the associated headlines and product titles. But the underlying, general sentiment rippling across the ecosystem is that you’re not entirely sure what it means for your org, your contract, your budget (flex crediiits 💪), and the ecosystem’s direction at large.

So, we set out to figure it out. What’s the right play?

This journey to demystification led to regular, 2-hour conversations with Geoff Momin from Interlay 💬 And after months of chewing it over, it feels right to open our notes to y’all 🙂

Should you be early adopters and plunge into all the world of AI has to offer?

Or take the wait and see approach with a risk of being left behind?

That’s what we cover in this issue today.

Let’s begin by orienting to some of those Workday “moves”.

Everything below is public information: press releases, keynotes, product pages, and a wind-down notice. (If you’re already on top of these happenings, scroll on.)

That’s a lot of moving pieces within a short span of time. So let’s assess…

If you focus on the Flowise thread of this story, you see that within 12 months a product was acquired, scheduled for go-live, renamed, absorbed into another acquisition, and then its open-source lineage completely wound down. It shows us how quickly one approach to building agents can experience pressure and fold.

If you look at the Sana story, you can see that its greatest impact so far has been as a new “face” of Workday. It updated the UI and delivered some encouraging (but not quite there) agents for those who’ve signed the UMSA and purchased Extend Pro.

If you focus on the developer story, you see something a bit different. Workday stopped asking you to come build in its new canvas and started meeting developers in Cursor, Codex, and VS Code — with agents reaching Workday data through MCP, under the security model you already run.

The canvas churned. The branding churned. The SKU it all lives in churned. But what held on was the stuff built on standards and security that already existed.

The long and short of it is…

None of these moves change your job as a Workday customer, which is to establish which capabilities fit the work, what they cost you, and who maintains them.

The data model, business processes, security groups, and audit obligations that any of these AI tools have to respect are the same ones you’ve had all along!

That's what a persistent layer looks like, and it's also where your playbook should live.

If you've got a proven use case and a tool that obviously clears your requirements with undeniable ROI — buy it.

But, if you're still working out the problem, let it simmer.

Waiting gives you time to get crystal clear about your problems, assess the stickiness of your vendor's product roadmap, and figure out best-fit tooling as you get intimate with your processes and what's possible. If your existing platform can do the job by then, you may save yourself another purchase and another integration.

It also gives your largest vendors (like Workday!) the time they need to pull their acquisitions and product plans together — fully. A year ago, the bet on Flowise would've been hedged against a product list that's since been renamed and fragmented twice. The cost of having waited on it? Zero.

And there are other hidden costs to blazing ahead.

Picture the same HR policy collection wired through Sana, Copilot, and Claude because you bought into each of these tools. To manage this responsibly, each connection needs a dedicated owner. Access needs checking in three places. Somebody has to investigate when an employee gets different answers depending on where they asked 🙃 Add on skills, actions, and separate admin consoles, and — wabam! — IT now has a fair amount of interface hell to look after with most of it integrated before anyone focused on a legitimate problem for it to solve.

At Rising, you’ll hear a lot of forward-looking statements. Ask the tough questions: what’s available now, what problems can I solve with it, what’ll it cost, and what would we miss by holding off for now?

Be slow. Be calculated.

But not in everything…

While vendors work through their products and messaging, get your team experimenting ASAP.

And we’re not talking about lame-o prompts like “Can you help me write this email…” or “Generate a plan for this project…”. Both of these can lobotomize thinking and churn out fluff that sounds good, says nothing, and usually gives the recipient a boat load of extra work as they sift through it 🙅🏻‍♀️

Instead, pick a problem — the one causing your org the most friction, that your team also knows well enough to catch a plausible-but-wrong answer to, is a great place to start.

Then, give your team room to build, and test their ideas with the people affected to learn whether it’s actually helpful.

For example, ask Claude or Copilot to draft a role-change checklist from a sample worker record and your relevant policies, using data approved for the exercise. Try a few cases and review the results with whoever normally does the work. See where the checklist needs correction and whether it saves the person reviewing it any time and effort.

Write down everything the exercise didn't handle.

Who should see the worker record? How do conflicting policies get resolved? How are mistakes caught and corrected? Who owns error handling, approvals, action logs, and support when the person who built it is on vacation? 🧐

Those gaps are your production requirements. They're also the most useful thing you can bring to a vendor conversation. Ask them to show you how they would handle the cases you tested, with your constraints. You may find a product already covers much of the work, or that a modest integration is enough to finish the job.

In this whole experimentation process, your team gets into the habit of questioning how the work gets done and where AI could have the most meaningful impact.

They also get the satisfaction of seeing something they’ve investigated (or built!) contribute to making someone’s work easier, giving you confidence to take on the next problem 😃

The names, the branding, and the roadmap will keep changing. But what won’t is your data model, your business processes, your security groups, and — most importantly — your team’s responsibility to gather criteria, design impactful solutions, and suss out the right tooling to deliver it.

Waiting to buy costs you nothing right now. Waiting to learn costs you the year.

In the next issue, we get you ready for those conversations — choosing a problem, testing an idea, and turning what you learn into requirements and questions for Workday and their partners.

See you there!

Did this issue help you out? Make you laugh? Did we miss any mission critical or helpful details?

Email us at [email protected] or leave a comment below — we'd love to know your thoughts 😊