Ashby Is Hiring 60 More Engineers While Greenhouse Cut 120. Source the Runners-Up.

The May 12, 2026 "Who Is Hiring?" thread on Hacker News is the clearest snapshot we've had of the recruiting tech bifurcation. Ashby posted that it grew engineering from 30 to 60 in 2025 and plans to hit 120 by year-end, at $153K-$220K for Senior and $190K-$275K for Staff Product Engineer, fully remote. Meanwhile Greenhouse and Lever's history (120 and 86 cut respectively in prior rounds) sets the priors for what happens to the incumbent ATS category when capital tightens.

If you run talent at an AI-native company, this asymmetry is your sourcing plan for the next 90 days. The narrow band of employers Ashby is competing with for senior ICs is the same band you're competing with, and the pool sitting between them (Ashby's cohort runners-up, displaced Big Tech recruiting ops, the Microsoft Rule of 70 cohort) is the most liquid it has been since 2023.

What Ashby actually disclosed

Ashby's HN post sits inside item 47975571, a thread with 321 listings, well below April's 400+. That compression matters: supply is concentrating into a smaller set of AI-native employers, and Ashby is the loudest voice in the room.

The numbers from their Series D announcement (July 2025, $50M led by Alkeon, ~2x the Series C valuation) tell you why they can spend on senior comp:

135%

Headcount has run from ~250 to roughly 320 across 2026, with engineering specifically slated to double from 60 to 120. The customer list (OpenAI, Shopify, Notion, Ramp, Cursor, Harvey.ai, Snowflake, Linear) reads as a who's-who of the same companies posting alongside Ashby on the same HN thread. Many of those listings link out to jobs.ashbyhq.com URLs. Ashby isn't just hiring inside the AI-native cohort, it's the ATS underneath most of it.

The "rejects" framing is wrong

The shortcut headline is "source Ashby's rejects." Don't run that play. Ashby publishes its hiring process on the team page and explicitly describes a cohort-based loop designed to compare candidates against each other rather than against a fixed bar. The output of that system isn't a pile of rejected mediocrities. It's a queue of objectively strong senior product engineers who lost a single slot to one other objectively strong senior product engineer.

These are runners-up, not rejects. The economic value of identifying them within 14 days of their final-round loss is enormous, because the median placement time for engineers with 8+ years of experience through specialized recruiters is 17 days, versus 45 to 90 days through generalist channels. The window closes fast.

Ashby's cohort process doesn't produce rejects. It produces runners-up who are 17 days from signing somewhere else. </pull> The practical sourcing move: identify the 30 to 50 engineers who reached final rounds at Ashby, Cursor, Harvey, Linear, and Notion in Q1 and Q2 2026 and didn't get an offer. They are public, traceable through GitHub activity timing, conference talks, and the "looking" signals you can read off LinkedIn and Polywork. They were vetted by the highest-bar interviewers in the category. You don't need to vet them again. You need to reach them first. This is the kind of search where natural-language sourcing tools earn their keep. Describing "senior backend engineers who interviewed at Ashby, Linear, or Harvey in the last 90 days, currently at a Series B+ company, with Rust or TypeScript depth" in a sentence is faster than building a Boolean string for it. That's the use case [Refolk](/) was built for: you ask in plain English, and the system pulls a ranked shortlist across GitHub, LinkedIn, and the open web. ## The bigger pool: Big Tech recruiting ops The non-obvious 2026 sourcing pool isn't displaced ATS engineers. It's displaced Big Tech recruiters. Meta confirmed earlier in this cycle that recruiting was the first function cut. Google trimmed recruiting alongside sales, product, and engineering. Amazon and Microsoft have ongoing reductions, and PayPal announced 4,760 cuts on May 9, 2026 as part of a $1.5B AI overhaul. The aggregate YTD layoff count from TrueUp sits at 128,270 across 286 events as of May 2026, roughly 1,002 people per day. Challenger reported 52,050 tech-sector cuts in Q1 2026 alone, the highest Q1 since 2023, +40% YoY.

stat number: 128,270 label: tech workers laid off YTD by mid-May 2026 note: 286 events tracked by TrueUp, roughly 1,002 people per day. </stat>

Inside that number sits a specific subpopulation: senior in-house TA leads who built and ran Greenhouse and Ashby implementations at companies with 5,000+ engineers. They know the workflows, the integrations, the report definitions. Hiring one of them to run your TA org is dramatically cheaper than hiring a Staff engineer, and the productivity ramp is immediate if you've standardized on Ashby (which, per the May HN thread, you increasingly have).

The mistake recruiting leaders make here is searching for "TA leader" as a title. The signal you want is "ran the Ashby or Greenhouse rollout at a 1,000+ engineer org, exited in the last 12 months." Title search misses 70% of that pool. The query needs to read profile content, not headers.

prompt: Senior in-house recruiters who led Ashby or Greenhouse rollouts at companies with 1,000+ engineers and left their role in the last 12 months
note: A ranked shortlist with current status, last-employer signal, and a permalink to the public profile evidence behind each match.
</refolk>

## The Microsoft Rule of 70 cohort

The most structured, time-boxed senior IC pool hitting the market in 2026 is Microsoft's voluntary retirement program. It's the first in the company's 51-year history. ~8,750 US employees qualify under the "Rule of 70" (age plus tenure). Notifications went out May 7. Decisions are due ~June 6.

This is not vague "layoffs are happening" wallpaper. It's a calendared event with a known cohort, known seniority distribution (you don't hit Rule of 70 at Microsoft as a junior IC), and a known decision window. If you read this in mid-May and your outreach campaign isn't dated for May 20 through June 13, you've already missed it.

The Microsoft cohort skews toward Principal and Partner-level engineers who've spent a decade-plus on Azure, M365, GitHub, or LinkedIn infra. Most aren't actually retiring. Most are using the package to fund a 12-month look at startup roles. They're not on LinkedIn job-seeker filters. They're not in your ATS. They show up in commit history, conference proceedings, internal Microsoft talks that leak to YouTube, and old patent filings.

### How to actually find them

The targeting heuristic that works: filter for Microsoft engineers with public commit activity slowing in March-April 2026 (the period between announcement and notification), public LinkedIn updates that quietly drop the Microsoft logo in early May, and a co-author or talk history with a known AI-native peer.

A natural language query like "engineers at Microsoft for 15+ years with Azure or GitHub experience whose public output dropped off in the last 60 days" is how you scope this pool in one sentence rather than 40 Boolean operators. That's again where Refolk's plain-English search pays off: the work of translating "Rule of 70 cohort" into queryable signals is the part that traditionally eats a sourcer's week.

## Don't blame AI for the incumbent collapse

A clean read of the data resists the lazy "AI ate recruiting tech" narrative. Challenger's own breakdown attributes only about 25% of March cuts to AI or automation. The other 75% is balance-sheet correction. Greenhouse and Lever started struggling well before generative AI shipped. The incumbent ATS category had pricing pressure, slow innovation cycles, and aggressive multi-product expansion that didn't pay back.

Ashby's edge isn't "we have AI." It's a burn multiple under 1x, 135% ARR growth, and a product wedge that started in analytics rather than applicant tracking. Gartner's projection that >60% of recruiting teams will use AI copilots by 2026 is a real tailwind, but it's the icing, not the cake. 55% of the 1,000 U.S. hiring managers Resume.org surveyed expect more layoffs in 2026; 44% cite AI as a top driver. That's hiring-manager perception, not the underlying P&L story.

The implication for sourcing: don't assume every laid-off Greenhouse or Lever engineer is a "stale" hire. The reason their company struggled has very little to do with their technical depth. The Ashby ATS is largely Ruby and TypeScript; Greenhouse and Lever talent maps cleanly into it.

## A 90-day sourcing plan from the May 2026 HN thread

If you took the May 12 thread as your only input, the right plan looks like this:

1. **Cohort runners-up.** Identify final-round candidates from Ashby, Cursor, Harvey.ai, Linear, Notion, and Synthesia over the trailing 90 days. Reach within 14 days of the rejection signal. These are senior product engineers in the $153K-$275K band who already cleared a top-decile bar.

2. **Displaced Big Tech recruiting ops.** Pull the Meta, Google, Amazon, and Microsoft in-house TA leads who shipped Ashby or Greenhouse rollouts. Title-blind search. Profile-content search. This is where you fill your TA function, not your engineering function.

3. **Microsoft Rule of 70.** Build the cohort list now. Outreach window: May 20 through June 13. After June 6, the people you want will have signed, and after June 13 the public signals will be too noisy to triage cleanly.

4. **AI-native peers.** Synthesia, Hyperspell, Chestnut, SentiLink, Atom Computing, Matterworks, Checkly, Comity are all in the same May HN thread. Their second-place candidates are your first-place candidates, and vice versa. Assume reciprocal poaching is already happening.

The thread compression to 321 listings from 400+ in April isn't a market slowdown. It's a clarity signal. The same ~30 employers are absorbing most of the senior IC supply, and they're using the same ATS to do it. If you're hiring out of the same pool, build your sourcing motion around the bifurcation, not against it.

## FAQ

### How do I find Ashby's final-round candidates without an inside source?

You triangulate. Look for senior engineers who appear in Ashby's own jobs.ashbyhq.com referer logs (public via some browser extensions), engineers who tweeted about interviewing at AI-native companies in Q1/Q2 2026, and public LinkedIn "open to work" signals dated 7-30 days after a high-bar AI-native company finished a hiring cohort. Cross-reference against GitHub activity for the Rust, TypeScript, or React profile Ashby specifically hires for. Tools that combine GitHub, LinkedIn, and open web in one search collapse this from a week of Boolean to one prompt.

### Are laid-off Greenhouse and Lever engineers actually worth sourcing?

Yes, but for different reasons than people assume. The technical work at incumbent ATS vendors is non-trivial multi-tenant infrastructure, integrations, and reporting at scale. The reason those companies struggled was pricing and product strategy, not engineer quality. They map cleanly onto AI-native recruiting roles, and they come at a discount to ex-FAANG comp expectations because they've been outside the FAANG comp curve.

### Is the Microsoft Rule of 70 window really only 14 days?

For the high-signal portion, yes. Notifications went out May 7. Decisions are due ~June 6. The two weeks after June 6 (~June 6 to June 20) are when the strongest candidates in that cohort are still evaluating offers. After that, you're competing with offers already on the table. The cohort doesn't fully clear the market until late Q3, but the top decile is gone by July.

### Should I just wait for inbound from these pools?

No. The 17-day median time-to-hire for engineers with 8+ years of experience through specialized recruiters versus 45-90 days through generalist channels is the entire game. Inbound is generalist-channel speed. Senior ICs in a bifurcated market get scooped before they ever surface in your ATS. If you can't query the open web in plain English and act on the result the same day, you're not in the running for the top of this pool.