Oracle's 10-K Just Dumped 8,000 Cerner Engineers Into Kansas City

Oracle's fiscal 2026 10-K, filed June 22, 2026, quietly did two things at once. It became the first S-1-level filing to name AI as a direct cause of workforce reduction, and it confirmed that 8,000 to 10,000 of the 21,000 net cuts came from Oracle Health, the entity Oracle built on top of its $28.3 billion Cerner acquisition. Almost all of that engineering depth sits in Kansas City, and it is on the open market right now.

If you run healthtech sourcing and you are not already working the KC metro this quarter, you are ceding the densest EHR engineering talent pool in the country to Cleveland Clinic, Kaiser Permanente, and a handful of staffing firms that have been building lists since April.

What the 10-K actually said

Oracle's headcount dropped from 162,000 to 141,000 over fiscal 2026, a net reduction of roughly 21,000 people, or about 13% of the workforce. Restructuring costs came in at $1.84 billion in severance and exit payments, compared to $374 million the year prior. TD Cowen estimates the cuts will free $8 to $10 billion in annual cash flow.

The line securities lawyers will be citing for years is this one: "the adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce." That is not an earnings-call throwaway. That is a regulatory disclosure. Cloudflare's Matthew Prince cut 20% and blamed AI in interviews. Snap's Evan Spiegel cut 1,000 citing "rapid advancements in artificial intelligence." Oracle just put it in the 10-K.

For founders pitching AI-native clinical operations, that changes what you are allowed to say out loud in a fundraising deck. For sourcers, it changes where the talent lives this month.

8,000-10,000

Why Kansas City specifically

When Oracle closed the Cerner deal in 2022, it inherited 28,000 Cerner employees, most of them anchored to two Kansas City campuses. Between 2022 and early 2026, Oracle had already trimmed roughly 5,000 KC jobs and closed two local offices. The March 2026 wave is the largest single event since the acquisition. WARN Act filings confirm 539 KC layoffs effective May 26 and June 1, but the true number is a multiple of that, spread across email-based terminations that did not trigger WARN thresholds.

On the morning of March 31, 2026, thousands of Oracle employees opened their inboxes to find a note from "Oracle Leadership" confirming their role had ended, effective immediately. For most of them, there was no prior conversation and no HR call. That matters for sourcing tone: these are not people who chose to leave, and they are not people who got a graceful runway. Cold outreach that acknowledges the abruptness lands very differently than a generic "saw you were open to work."

Oracle Health is now headquartered in Austin, with intentions to move to Nashville. The company that built KC's digital health identity has functionally left. That has two effects. First, displaced engineers are unusually willing to consider remote roles and relocation packages. Second, KC's ecosystem, roughly 25,000 people working in digital health, including 6,000+ with 10+ years of experience and 2,000+ with 20+ years, is not going to be re-absorbed by Oracle. It has to go somewhere.

The pipeline is still running

Johnson County Community College's Health Information Systems program trains more than 500 students a year, and they get hands-on Oracle Cerner EHR experience as part of the curriculum. KU runs CAHIIM-accredited Health Information Management and Informatics programs. UMKC has its own pipeline. This is not a one-shot supply event. KC keeps producing Cerner-fluent graduates whether Oracle Health is headquartered there or not.

CBRE's 2025 Scoring Tech Talent Report ranked KC as the 6th most cost-effective tech market in the U.S. If you are hiring remote-first from a coastal budget, KC salaries stretch further than almost anywhere else you would seriously consider for domain talent.

What "Cerner engineer" actually means

This is where most generalist recruiters get it wrong. Cerner engineers are not interchangeable with Epic engineers or with generic backend engineers who happen to have worked on healthcare data. The Cerner stack is its own world.

• CCL (Cerner Command Language): Millennium's proprietary query and reporting language. Almost nobody outside the Cerner ecosystem has touched it.
• PowerChart: The clinician-facing workflow shell. Customization work here is a distinct skill.
• HL7 v2 interfaces: Legacy point-to-point messaging that Epic-native engineers often skipped entirely because Epic pushed FHIR earlier.
• FHIR: Modern standard. Cerner engineers who did integration work in the last five years usually have this, but it is not universal.
• Millennium architecture: The database and application layer that underpins everything else.

The titles you are looking for in the wild are Clinical Informatics Analyst, EHR Integration Specialist, Senior Interoperability Specialist, and Healthcare Data Engineer. Those are not job families most general-purpose sourcing tools index well. They are also not roles where a boolean string like "Cerner" AND "HL7" gets you a clean list, because a lot of the deepest people have titles that never mention Cerner explicitly (they were just at Cerner, then Oracle, doing the work).

This is one of the reasons we built Refolk: you describe the person in plain English ("Cerner integration engineers in the KC metro who know CCL and HL7 v2, laid off from Oracle in the last six months") and get a ranked shortlist without hand-crafting boolean strings against titles that do not map cleanly.

The window is measured in weeks

Here is the part that should make you move this week rather than next quarter. A Cerner integration engineer we tracked was placed into a regional health system 11 days after the March 31 layoff. By day 14, that same candidate already had two other conversations converting into formal offers.

The "cheap displaced talent" narrative assumes a slow-moving market. That is not what is happening here. HPG Resources, Kore1, ClinDCast, and IMO Tech Solutions have been running lists since April. Cleveland Clinic is the single largest employer of Cerner/EHR/HL7/FHIR-skilled specialists in the professional index, and it recruits nationally. Kaiser Permanente is active. Every remaining Cerner-shop academic medical center is quietly pulling from this pool because they know their vendor is under structural pressure.

Oracle has lost 57 acute care clients in three years. Every one of those switching customers still needs Cerner-fluent staff to run parallel systems and manage the migration. The demand side is not going away, which means the good people get absorbed fast.

The company that built KC's digital health identity has functionally left. The engineers have not.

VA and DoD clearances are a bonus most people miss

Hundreds of Oracle Health employees worked on the VA EHR Modernization Program, especially after the government implemented financial penalties in 2023 for missed milestones. A meaningful slice of the displaced KC engineers hold active DoD or VA clearances. If you are building anything that touches federal health infrastructure, or defense health, or Medicare data pipelines, clearance-holding EHR engineers are a category so scarce that clearance alone justifies a relocation package.

Filter for it explicitly. Most sourcers will not.

How to work the pool without burning it

A few tactical notes, because the difference between doing this well and doing it badly is stark.

One, do not lead with "saw you were laid off." These people got fired by email on a Tuesday morning. They know. Lead with the specific problem you are solving and the specific skill you need. Cerner engineers respond to founders who understand the difference between an ADT feed and an ORM message. They ignore recruiters who do not.

Two, look at GitHub for the integration engineers. A surprising number of Cerner interoperability people maintain personal HL7 parsing libraries, FHIR sandbox projects, or Mirth Connect channels on GitHub. That is your signal for who actually writes code versus who project-manages other people who do. Refolk pulls GitHub, LinkedIn, and open-web signals together in one query, which matters here because the LinkedIn-only view of this population undersells the technical depth by about half.

Three, respect the KC network effect. This is a small industry city. Ex-Cerner people talk to each other constantly. If you burn one candidate with a lowball offer or a bait-and-switch job description, you will hear about it on the next three calls. Move fast, but move honestly.

Four, use the ecosystem organizations. OneKC / Digital Health KC, KC Tech Council, BioNexus KC, and the Keystone Innovation District all run programming that displaced Cerner engineers attend. Sponsoring a meetup is cheaper and higher-signal than a LinkedIn Recruiter seat.

Who should be moving right now

Three founder profiles should treat this as a five-alarm sourcing event.

Health system CIOs still running Cerner. You have 12 to 24 months before Oracle's product roadmap forces a hard decision. Hire the people who built it, before the Epic migration consultancies do.

AI healthtech founders building on EHR data. You need engineers who understand what the data actually means when it comes out of a Millennium instance. General ML engineers can build the model. Cerner engineers know which HL7 segment is lying to you.

Interoperability startups. Redox, Health Gorilla, Particle, and their competitors have been sourcing this pool for years. The March 2026 wave is the first time it has been abundant instead of scarce.

If you are not one of those three, you probably do not need to move this week. If you are, the KC candidate pool that exists on July 15 will not be the KC candidate pool that exists on October 15. Seema Verma can announce a new acute care EHR for 2026 all she wants. The engineers who would have built it are already taking calls from Cleveland.

FAQ

How do I verify a candidate was actually part of the March 2026 Oracle Health cut versus an earlier wave?

Ask for the effective date on their separation notice, and cross-check against LinkedIn "open to work" timestamps and any GitHub activity gaps. The March 31 cohort has a very specific pattern: activity stops abruptly on a Tuesday, LinkedIn headlines change within two to five days, and the tenure lists Oracle from 2022 (post-Cerner-close) or earlier if they were Cerner originals. WARN filings for KC name 539 people specifically, so those exact names are public record.

Are Cerner engineers a fit for a generalist AI startup, or only for EHR-adjacent work?

Only for EHR-adjacent work, in most cases. The domain depth is the whole point. If your startup does not touch clinical data, HL7, FHIR, or health system workflows, you are better off hiring generalist backend engineers and paying market rate. Cerner engineers know things that are extremely valuable in a narrow context and not especially transferable outside it. Do not hire them for a role that does not use what makes them rare.

What salary should I anchor to for a senior Cerner integration engineer in KC?

KC's cost-of-living-adjusted market rate is meaningfully below Bay Area or NYC, and CBRE ranks it 6th most cost-effective nationally. That said, do not lowball. The floor for a senior interoperability engineer with 10+ years of Cerner tenure and active FHIR work is competitive with regional health system director-level pay, because that is who is bidding. Match it, or offer remote flexibility and equity to close the gap.

Where does Refolk fit into an EHR sourcing workflow?

Refolk is the plain-English query layer on top of LinkedIn, GitHub, and the open web. You describe the person you actually want ("Cerner engineers in Kansas City who left Oracle in 2026 and hold VA clearances") instead of hand-building boolean strings against title fields that do not capture the real skill graph. It is most useful in exactly the kind of niche, high-domain-depth sourcing this article is about, where the difference between a generic search and a precise one is the difference between 40,000 noisy results and 200 real candidates.