Workday Cut 525 in Product and Tech. Most 'Workday' Profiles Just Use It

Workday's Sept 29 cut puts about 525 jobs on the table, most of them in Product and Technology, and the 142 in Pleasanton end on Nov 30. If you want to hire former Workday engineers, your first instinct is to search the employer name. It mostly fails, because "Workday" on a profile almost always means the person uses the software. It rarely means they built it.

What Workday cut on Sept 29, and when people are free

Workday is cutting about 2.5% of its workforce, roughly 525 jobs, "primarily within Workday's Product and Technology team," with $65M to $85M in charges. It follows an earlier 2026 round on April 6 that cut 154 jobs. The 142 roles at its Pleasanton headquarters end on Nov 30.

An 8-K is the filing a public company makes with the SEC to disclose a material event. A WARN notice is the advance filing a state requires before a mass layoff. In California it lists the site, the headcount and the effective date. Here is the record:

• Sept 29, 2026 (8-K): about 2.5% of staff, mostly Product and Technology, plus "select leased office space reductions."
• Sept 29, 2026 (California WARN): 142 permanent layoffs at 6110 Stoneridge Mall Road, Pleasanton, effective Nov 30. That is 62 days of notice.
• April 6, 2026: an earlier round that cut 154 jobs.
• 2025: a WARN notice for 617 jobs in Pleasanton.
• Still hiring: Workday says it "plans to continue to hire in key strategic areas and locations throughout its fiscal 2027."

Recruiters should care about this round because of where it lands: in the team that writes the product. One caution: the public WARN summary "does not identify individual job titles or establish a complete worldwide location breakdown." You know roughly how many people are leaving and where some of them sit. You don't know who they are. The rest of this piece is about finding them.

Why searching "Workday" returns its customers

Searching "Workday" fails because it is both an employer and a product that HR, finance and recruiting staff use every day. Users outnumber builders by orders of magnitude. In Refolk's index of professional profiles, 71,827 US profiles list "Workday" as a skill, and none of a 25-profile sample had ever worked there.

71,827

The reason is simple. A skill tag records what someone touches at work, and every company that runs Workday turns its employees into "Workday" profiles.

Narrowing to job titles helps less than you'd expect:

Two ratios explain the problem:

• Skill-tagged profiles outnumber engineer-titled ones about 156 to 1.
• Workday Consultant, Workday Analyst, HRIS Analyst and Workday Integration titles outnumber engineer titles about 3.3 to 1.

The biggest single employer in the consultant sample was Walmart, with 8 of 25. Walmart is a customer, and one large customer can bury the builders you're looking for. A sourcer who starts from "Workday" and pages through results screens HRIS analysts long before reaching the first engineer who wrote the code. Meanwhile, the people actually affected by the Workday software engineer layoffs are taking calls from someone else.

The search that works: engineering title plus "workday"

The best simple search starts with a software engineering title and then requires "workday" in the headline. In Refolk's index that returns 461 US profiles, and 18 of a 25-profile sample actually worked at Workday.

That is about 72%, which suggests roughly 330 of the 461 are real Workday engineers. Treat the number as directional, since it comes from a sample of 25. It is still a big improvement on 0 of 25.

The order of the filters matters. A title says what someone does, and an employer word in the headline only says what they work near. When you lead with the title, the product name mostly points back to the company. When you lead with the product name, the title has to fight off tens of thousands of users.

The 28% that slip through follow a pattern. They came from ADP, KeyBank, Dell, Lowe's, Standard Chartered and Meta. At those companies "Workday" means integration work. Check the employer field in their work history, not the headline, and they're easy to drop.

This is the general answer to how to source engineers from a specific company whose name is also a product. Start from the function, then confirm the employer from work history. A boolean string can do that, but it breaks as soon as someone writes "SDE II @ WD" or leaves the headline blank. Refolk handles that step differently. You describe the person ("software engineers who worked at Workday, not Workday customers") and get back a ranked list checked against employment history, not keyword matches.

Extend and Prism are consultant words, not builder words

Workday Extend and Prism Analytics look like insider terms, but they find consultants. Both are products Workday sells, partner consultants get certified in them, and searching them brings back the crowd you're trying to avoid.

The evidence is in job ads and partner directories:

• Consulting job ads ask for an "Active Workday Partner certification in Prism Analytics."
• A whole tier of implementation partners focuses on "Extend, Workday Studio, Prism Analytics, and API orchestration." Firms such as Invisors, TopBloc and Sability build their practices around these products.

So use Extend and Prism as exclusions. Add NOT (Prism OR Extend OR certified OR certification) to an engineering search to strip out partner profiles built around those certifications. Consultants list these products because that is what they're hired for, which is exactly why the terms can't separate builders from everyone else.

XpressO and OMS leak too, just less. XpressO is Workday's metadata layer, and OMS provides the runtime environment for it. Both sound internal. But training sites written for consultants explain them in detail, for example: "OMS also provides a runtime environment for XpressO." Architects studying for certification pick up the vocabulary and add it to their profiles.

A term stops being an insider signal the day a certification course starts teaching it.

The stack signals that separate builders

Workday's builders show up as an engineering title plus the languages Workday actually builds in, with XpressO or OMS as a bonus signal, never a filter on its own. Workday's own engineering blog says: "At Workday we use a variety of languages including Java, Scala, Python and Ruby."

Its platform whitepaper adds that "Workday's foundation is built on Java" and names XpressO as the metadata layer. That gives you a stack fingerprint to pair with the title. XpressO and OMS on their own pull in certified architects. Paired with an engineering title and Scala or Java, they point at builders.

Stack the signals in this order:

1. Engineering title. Software Engineer, SDE, Senior, Staff or Principal Engineer.
2. Workday in work history, confirmed as the employer, not mentioned in a description.
3. Scala or Java, the strongest backend signal.
4. XpressO or OMS, only as a tiebreaker on profiles that already pass steps 1 to 3.
5. Exclusions: Prism, Extend, certification, HRIS, Consultant.

Outside the profile, the Workday Engineering blog on Medium adds names: its bylines belong to working engineers.

Workday's GitHub org: read commits, not forks

Workday's GitHub org is a real insider signal if you look at who commits and opens pull requests, not who forks. The org has 55 repositories. The flagship, canvas-kit, has about 364 stars and 271 forks.

canvas-kit is a TypeScript UI kit for the Workday Sana Canvas Design System, developed in the open. Its contributor list is public at github.com/Workday/canvas-kit. The sibling repos, canvas-tokens and canvas-icons, work the same way.

How to use them:

• Ignore the fork list. Many of the 271 forks come from unrelated personal accounts.
• Pull commit and PR authors. Commit and PR authors on Workday-owned repos are the insider signal, especially on a design system that ships Workday's own product UI.
• Cross-check against a profile. A GitHub handle plus a matching Workday employment record is about as strong as sourcing evidence gets.
• Know the skew. This pool leans toward frontend and TypeScript engineers, not the Scala and OMS backend teams, so you still need the stack search above to find them.

Why the Pleasanton radius covers only 27% of the cut

A commute-radius search around Pleasanton finds the most urgent group, not the largest one. The 142 WARN roles are about 27% of the roughly 525 total, so about 383 affected people (73%) work outside that notice.

73%

The Pleasanton group is still worth working first. Their Nov 30 date is fixed and public, they're within commuting distance of the Bay Area market, and California's notice requirement means you know exactly when they're free. Workday has cut in Pleasanton before (617 in 2025), so the East Bay already holds former Workday staff from earlier rounds.

For everyone else, location doesn't work as a filter. The remote engineers and people in other offices only show up through the stack and GitHub signals. Run both searches side by side: a radius search for speed, and a stack search for coverage. Then remove duplicates.

When to reach out: before Nov 30, not after

Reach out now, during the notice period, because affected engineers still show as employed at Workday until Nov 30 and won't appear in any "open to work" filter. Waiting for the signal means arriving after other recruiters.

Two things make timing tight:

1. The notice period hides people. With 62 days' notice, profiles still say "Workday" and current. Status filters miss everyone in this cut.
2. Internal moves take some of them. Workday says it will keep hiring "in key strategic areas." Some affected engineers will move into those roles before their exit date.

Write outreach that shows you know the difference between a builder and a user. Mention their actual stack or their canvas-kit commits. A message that could only have been written to an engineer stands apart from the ones written for HRIS analysts.

FAQ

How many people did Workday lay off in 2026?

The Sept 29 8-K announced about 2.5% of the workforce, roughly 525 jobs, mostly in Product and Technology, with $65M to $85M in charges. An earlier round on April 6 cut 154 jobs. A 2025 WARN notice separately covered 617 jobs in Pleasanton.

When do the Workday layoffs in Pleasanton take effect?

The California WARN notice dated Sept 29, 2026 lists 142 permanent layoffs at 6110 Stoneridge Mall Road, Pleasanton, effective Nov 30, 2026. That gives 62 days of notice, so affected engineers will likely still show as employed at Workday until the end of November.

Are Workday Extend or Prism good search terms for finding Workday engineers?

No. Extend and Prism Analytics are products Workday sells, and partner consultants get certified in them, so searching on them returns implementation partners and customer-side analysts. Use them as exclusions instead. XpressO and OMS are better, but only as a tiebreaker on profiles that already have an engineering title and Scala or Java.

How do you source engineers from a specific company when its name is also a product?

Start from the function, not the company name. Filter for engineering titles first, then confirm the employer from work history rather than headline keywords, then add the company's actual stack and public code contributions. For Workday, that moved the sample from 0 of 25 real employees (skill search) to 18 of 25 (engineer title plus "workday") in Refolk's index.