Refolk
August 25, 2026·8 min read

Apple's Aug 21 Cuts: 15 "Siri" Profiles Hide a Pool of 658

Apple's Aug 21, 2026 layoffs hit Intelligent Systems Experience, not just Siri. Here's how to source the real ex-Apple AI pool before others do.

Apple Intelligent Systems Experience layoffsex-Apple AI engineers sourcingApple Siri team layoffs 2026Vision Pro engineer poachISE Apple recruiter
Apple's Aug 21 Cuts: 15 "Siri" Profiles Hide a Pool of 658

Every recruiter running "ex-Apple" AND ("Siri" OR "Vision Pro") this week is fishing the wrong pond. On Aug 21, 2026, Apple confirmed cuts to more than 200 people, split roughly 100 in Vision Pro and 100 across Siri, Software Engineering, and a sub-org called Intelligent Systems Experience, per Bloomberg.

The real displaced AI talent sits inside Intelligent Systems Experience (ISE), the team responsible for on-device AI integration on Apple devices. Nobody puts "ISE" in a LinkedIn headline. Here's how the pool actually breaks down, and how to reach it before the other 40 recruiters figure out the trick.

Why "Siri" is the wrong keyword

Searching "Siri" at Apple returns about 2% of the relevant AI engineering pool. In Refolk's index, Apple US employees with "Siri" in their headline or keywords number 15 people total. Apple US engineers with ML, AI, or Research Engineer titles number 658. That is a 1:44 ratio.

The mismatch is structural. Apple engineers describe what they built, not which internal team they sat on. An engineer who spent two years shipping the App Intents integration for the new Siri AI is more likely to write "on-device LLM orchestration" than "Siri team." Internal org names like ISE rarely surface on public profiles, and Apple's culture actively discourages it.

Which means the "Apple Siri team layoffs 2026" search that half the industry is running right now is fishing in a 15-person pond next to a 658-person lake.

658
Apple US engineers with ML/AI/Research titles
In Refolk's index, only 15 of them have "Siri" anywhere in their profile.

What ISE actually is

Intelligent Systems Experience is a sub-org inside Apple's Software Engineering division responsible for the AI-powered capabilities that stitch together the new Siri. Bloomberg's Mark Gurman named it explicitly in his Aug 21 report, which is the first time most of the outside world had heard the label.

ISE owns the surfaces where models meet the OS:

  • Apple Foundation Models, the on-device LLM stack
  • Private Cloud Compute, the server-side inference layer for queries too heavy for on-device
  • App Intents and App Toolbox, the developer surfaces that let third-party apps expose actions to Siri
  • Spotlight's index, used as the personal-context retrieval layer
  • The system orchestrator that routes a user request across all of the above

If you are trying to hire someone who has actually built LLM orchestration in production at consumer scale, this is the team. The June 8, 2026 developer preview of Siri AI and the late-2026 public beta were shipped by these people.

The Vision Pro half is not what recruiters think

Vision Pro cuts are not a spatial-compute talent windfall. Apple is largely shutting down a Vision Pro team focused on gaming and reducing the immersive video content unit, per Engadget's Aug 21 reporting. That means most of the ~100 Vision Pro departures are content producers, Unity gameplay engineers, and video pipeline people, not visionOS kernel or ARKit engineers.

Refolk's index shows only 7 Apple US staff with "Vision Pro" or "visionOS" as a keyword in their profile. Combined with an earlier ~60-person Vision Pro cut reported by 9to5mac, cumulative Vision Pro attrition since early 2026 is closer to 160 people. But the core spatial-compute engineers, the ones an AR startup would actually want to poach, are largely still inside.

If your pitch is "come poach ex-Vision Pro engineers for our AR startup," you will mostly land on video producers. A Vision Pro engineer poach that yields spatial-compute talent requires filtering hard on ARKit, RealityKit, visionOS SDK, or compositor-level graphics experience, not the Vision Pro label itself.

The real numbers, side by side

Here is the dataset a sourcer should keep open in a second tab this week.

SegmentApple US countNotes
ML/AI/Research Engineer titles at Apple658The actual AI pool
Profiles with "Siri" keyword1510 SWE, 4 MLE, 1 SWE Manager
ML Engineers within the "Siri" slice4The narrow "obvious" pool
Profiles with "Vision Pro" or "visionOS" keyword7Most cuts are gaming/video, not these
Announced cuts across Siri + SWE + ISE (Aug 21)~100Bloomberg via Yahoo Finance
Announced Vision Pro cuts (Aug 21)~100Mostly gaming and immersive video

Source for the count rows: Refolk's index of professional profiles, US-filtered.

The interesting derived number is that the "Siri"-labeled pool is 1:44 of the real Apple AI pool. If you are only Boolean-searching "Siri," you are competing hard for 2% of the market while ignoring 98%.

How to identify ISE alumni by proxy

You cannot search "ISE" and get anything useful. You have to proxy via the artifacts ISE ships. Here is the keyword stack that actually returns the right people:

  1. Foundation Models (Apple's specific term for its on-device LLM family)
  2. Private Cloud Compute or PCC
  3. App Intents and App Toolbox
  4. iOS 27 paired with AI/ML or "assistant"
  5. On-device inference, on-device LLM, model orchestration
  6. Spotlight paired with retrieval, embeddings, or ranking
  7. System orchestrator or query router

Any Apple engineer whose profile mentions two or more of these terms is almost certainly ISE or ISE-adjacent. Any engineer who mentions Foundation Models plus Private Cloud Compute is a near lock.

This is exactly the search problem where Boolean falls apart and plain English matters. You want "current or recently former Apple engineers in the US who shipped on-device LLM features for the new Siri, ranked by seniority." Writing that as a keyword expression means enumerating a dozen synonyms and hoping. That is the exact gap Refolk closes: you describe the person in plain English and get a ranked shortlist across GitHub, LinkedIn, and the open web, without having to guess every internal Apple codename.

The 6-week timing window

Engineers cut on Aug 21 almost certainly shipped the June 8 developer preview, which puts a hard clock on this pool. Apple opened Siri AI to developers on June 8, 2026, with public beta planned for late 2026. Anyone laid off in August was still on the team through the developer preview ship. That is production LLM orchestration experience at consumer scale, which almost nobody outside Apple, Google, and Anthropic actually has.

The window to reach them warm is roughly 6 weeks. After that, Meta Reality Labs, Google DeepMind, Anthropic, and OpenAI recruiters will have closed most of the senior conversations.

Anyone laid off in August almost certainly shipped the June developer preview. That experience does not exist anywhere else on the market.

Two operational notes on the window:

  • Reshuffles absorb 30 to 50%. Apple's official statement said it will "create new roles as part of this change." Some ISE engineers are being moved into the new Siri AI org, not fired. Assume the truly available external pool is 50 to 70 people, not 100.
  • The Vision Pro overlap is real but small. A handful of ISE engineers likely rotated through Vision Pro for on-device model work on visionOS. If you see both surfaces in the same profile, that person is a top-decile target.

Why "ex-Apple SWE" is a useless tag

The Apple SWE org contains ISE, Siri legacy, Core OS, iCloud, Services, and dozens of other teams, so "ex-Apple SWE" is close to meaningless as a segmentation. The 658 vs 15 gap in Refolk's index is the cleanest illustration of why title and employer sourcing breaks down at Apple specifically.

At Apple, internal identity is deliberately opaque. Engineers describe their public-facing work: the framework they contributed to, the WWDC session they presented, the iOS version they shipped in. That is the actual sourcing surface.

Which is why ex-Apple AI engineers sourcing works best when you index on artifacts rather than org labels. Set your filter to WWDC 2026 presenters, Apple ML Research paper authors, and profiles mentioning iOS 27 features. Cross that with a departure signal (updated profile in the last 30 days, "open to work," or a status change) and you will get a cleaner list than any "ex-Apple Siri" search will produce.

A concrete sourcing playbook for this week

Run three parallel searches, not one, and prioritize the ISE proxy stack over the Siri keyword.

  1. Search 1: ISE proxy. Apple US, ML/AI/Research titles, mentions of Foundation Models OR Private Cloud Compute OR App Intents. Expect a few hundred results. This is the core pool.
  2. Search 2: Departure signal overlay. Same filter as Search 1, plus profile updated in the last 45 days or explicit "open to work." This is your outreach list for the next 6 weeks.
  3. Search 3: Vision Pro spatial only. Filter hard on ARKit, RealityKit, or visionOS SDK contributions. Ignore anything tagged only "Vision Pro." Expect a very small list, well under 20 nationally.

The rest of the market is spending this week InMailing 15 people. The pool you actually want is 40 times larger and hides behind three keywords Apple's marketing team invented.

FAQ

What is Intelligent Systems Experience at Apple?

Intelligent Systems Experience, or ISE, is a sub-org inside Apple's Software Engineering division responsible for on-device AI integration, including the new LLM-based Siri. ISE owns the Apple Foundation Models runtime, Private Cloud Compute, App Intents, the system orchestrator, and the Spotlight-backed personal context layer. It is not a marketing brand, so engineers on the team almost never mention it on public profiles.

How many Apple engineers were actually laid off on Aug 21, 2026?

Apple confirmed more than 200 cuts, split roughly 100 in Vision Pro and 100 across Siri, Software Engineering, and ISE, according to Bloomberg's Mark Gurman. The Siri cuts specifically targeted the team maintaining the legacy assistant, because the new LLM-based Siri requires different expertise. Vision Pro cuts were concentrated in gaming and immersive video content teams, not core visionOS engineering.

Why doesn't searching "Siri" find the right ex-Apple AI engineers?

In Refolk's index, only 15 Apple US employees have "Siri" in their profile out of 658 with ML, AI, or Research Engineer titles. Apple's internal team names rarely appear on LinkedIn, so the "Siri" keyword returns roughly 2% of the relevant pool. Proxy keywords like Foundation Models, Private Cloud Compute, App Intents, and iOS 27 return far more of the actual ISE and ISE-adjacent talent.

How long is the window to reach these engineers before they sign elsewhere?

Roughly 6 weeks from the Aug 21 announcement. Apple shipped the Siri AI developer preview on June 8, 2026, so anyone cut in August has fresh production LLM orchestration experience, which Meta Reality Labs, Google DeepMind, Anthropic, and OpenAI are all hunting. Assume 30 to 50% get reshuffled into new internal Apple roles and never hit the market, leaving a real external pool closer to 50 to 70 senior engineers.

Try it on the search you came here for

Stop building boolean strings. Just describe the person.

Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.

  1. 01Describe them

    One plain sentence. Role, city, stack, stage, whatever matters to you.

  2. 02I read the web live

    GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.

  3. 03You read the shortlist

    Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.

  • No boolean, no filters, no seat to buy. One box.
  • Read at search time, so a profile updated yesterday counts today.
  • Every step visible as it runs, every name with its reason.

500 free credits on sign-up. No card, no demo call. See real searches.

Read next