# Yelp's First Layoff Since 2020: Source Its Engineers From PaaSTA Commits

_Yelp's Nov 18 layoff hits nearly 5% of staff. A 'Yelp' keyword search mostly finds US alumni. Source its infra engineers from recent PaaSTA commits instead._

- Canonical URL: https://www.refolk.ai/blog/yelp-layoffs-source-engineers-paasta-github-commits
- Published: 2026-11-20
- Reading time: 10 min
- Keywords: Yelp layoffs November 2026, Yelp engineers hiring, source engineers from GitHub contributors, Yelp PaaSTA open source

If you search for Yelp engineers this week, the obvious query - "Yelp" in the headline - mostly returns people who left years ago. In Refolk's index, only 2 of the top 25 US software engineers with "Yelp" in their headline still work there.

Yelp told employees on November 18, 2026 that it is cutting nearly 5% of its full-time staff. It is the company's first workforce reduction since the 2020 pandemic layoff, and every recruiter with a platform req will be working the same list. Yelp's public GitHub org gives you a more accurate map of who builds its infrastructure today.

## What did Yelp announce in its November 2026 layoff?

Yelp filed an 8-K saying that on November 18, 2026 it notified employees of a workforce reduction "expected to impact nearly 5% of our full-time employees," with charges of about $14 million to $16 million. It did not give a headcount or name the teams.

AIM Group reports about 250 employees. That figure comes from applying 5% to the headcount in Yelp's 10-K, and Yelp has not confirmed it. Yelp also has not said which departments or locations are affected. So nothing public says engineers were cut. You are sourcing from a company in transition, not from a known list of laid-off engineers.

Compared with 2020, this cut is small:

| | 2020 | November 2026 |
|---|---|---|
| People cut | 1,000 (plus 1,100 furloughed) | About 250 (estimated, unconfirmed) |
| Share of staff | 17% of 5,950 | Nearly 5% |
| Stated reason | Pandemic | Focus on AI-powered offerings in the "other revenue" line |
| Restructuring charges | Not cited | $14M to $16M, Q4 2026 to Q1 2027 |

The financial backdrop is tight. Q3 2026 net revenue fell 1% year over year to $371 million. At the end of June, Yelp held $94 million in cash and cash equivalents against $100 million in net borrowings. The 8-K says Yelp plans to focus on opportunities "including the AI-powered offerings in our other revenue line." Read the cut as a reallocation: leadership under co-founder and CEO Jeremy Stoppelman is moving money toward AI, which matters for who you should pitch (more on that below).

## Why does a "Yelp" keyword search mostly find alumni?

In the US, a headline search for "Yelp" mostly finds people who no longer work there. In Refolk's index of professional profiles, 217 US profiles combine a software engineer title with "Yelp" in the headline. In the top-25 sample, only 2 list Yelp as their current employer. Meta also shows up 2 times.

> **2 of 25** US "Yelp" software engineer profiles that currently work at Yelp
>
> In Refolk's index, the other 23 in the top-25 sample have already moved on.

Two things cause this:

- **Headlines outlast jobs.** Engineers keep "ex-Yelp" in their headline for years because it still helps them. A keyword match can't tell "Senior Engineer at Yelp" from "ex-Yelp, now at Meta."
- **2020 created a large alumni pool.** Cutting 1,000 people in one round put a lot of Yelp-branded profiles on the market at once.

Here is the full picture from the index:

| Segment | Profiles with "Yelp" in headline | Currently at Yelp (top-25 sample) | Top regions |
|---|---|---|---|
| SWE / Senior / Staff, US | 217 | 2 of 25 (8%) | NYC, SF, Bellevue |
| SWE / Senior / Staff, UK + DE | 125 | 18 of 25 (72%) | London, Berlin |
| SRE / Platform / Infra / DevOps, US + UK + DE | 39 | 9 of 25 (36%) | Mannheim, LA, London |

The current-employer shares come from 25-profile samples, so read them as a direction, not a census. Even so, the gap between 8% and 72% is wide enough to change how you search in each country.

If you are sourcing Yelp engineers in the US, filter on current employer, not on a headline keyword. This is where [Refolk](/) helps: you can ask for "engineers currently at Yelp in New York" in plain English and get back people whose current role matches, not everyone who ever mentioned the company.

## Why does the same search work in London and Berlin?

In the UK and Germany, a "Yelp" search mostly returns current staff: 18 of the top 25 among 125 profiles. The same query that fails in the US works reasonably well in Europe.

The index doesn't explain why, but the likeliest reading is the US problem in reverse: fewer "ex-Yelp" headlines diluting the results. The US has 1.74 times as many matching profiles, yet the European list has a far higher share of current employees.

Two cautions:

1. Yelp has not said whether London or Berlin were affected. Treat European staff as people who may be open to a conversation, not as people who need a job.
2. A clean list of current staff in London is also the list every other recruiter in London is working this week. Speed matters less than a better opening message.

## How do you source engineers from Yelp's GitHub contributors?

Start with a repository, identify the team that owns it, and then filter contributors to recent releases instead of all-time rankings. Yelp's GitHub org has 102 public repositories, and several of them map to specific teams.

| Repo | Stars / activity | Team signal |
|---|---|---|
| PaaSTA | Shipped v1.66.1 on Nov 19, 2026 | OWNERS file names Compute Infra |
| detect-secrets | About 4.6k stars | Maintained by Yelp's security team |
| dumb-init | About 7.3k stars | Most-starred, but stars don't tell you the team |
| mrjob | About 2.6k stars | Data infrastructure |
| Tron | Last updated Aug 27, 2026 | Data infrastructure |

PaaSTA is the best place to start because its OWNERS file names exactly one team: Compute Infra <compute-infra@yelp.com>. That is the closest thing to an org chart Yelp publishes.

The trap is the contributor leaderboard. All-time lists favor the engineers who built the project, not the ones maintaining it now. On libraries.io, PaaSTA's top all-time committer, Kyle Anderson, holds 946 commits on his own, and early names like Tyler Roscoe and Evan Krall also rank high. That is project history, not a current roster.

> **27.2%** Share of all PaaSTA commits held by its single top all-time contributor
>
> All-time rankings measure who built the project, not who maintains it now.

To source engineers from GitHub contributors without ending up with an alumni list:

1. **Pick repos by team, not by stars.** PaaSTA for Compute Infra, detect-secrets for security, mrjob and Tron for data infra.
2. **Read release notes from the last 6 to 12 months.** The September to November 2026 PaaSTA releases show who is actively shipping.
3. **Check commit authors on the default branch** in that same window, and drop bots and one-off drive-by contributors.
4. **Match each handle to a professional profile** to confirm location, title, and current employer.
5. **Note which team they belong to** so your pitch addresses their actual work.

Step 4 takes the most time when you do it by hand, because GitHub handles rarely match names and many committers have no bio. I built Refolk to search GitHub, professional profiles, and the open web together, so you can start from a commit and get back the person behind it.

> **Example search:** "Find engineers who contributed to Yelp's PaaSTA, Tron, or detect-secrets repos in the last 12 months, and show their current title and location." - [run it on Refolk](https://www.refolk.ai/s/hqzgzng8a7)
>
> You get a ranked shortlist of recent committers matched to their professional profiles, with alumni separated from current staff.

## Does a stopped commit history mean someone was laid off?

No. A gap in someone's commits is not evidence of a layoff, and using it that way is both unreliable and creepy. PaaSTA shipped v1.66.1 on November 19, the day after the announcement, and that release included a first-time contributor. The repo kept moving.

Commits stop for ordinary reasons:

- Vacation or leave
- A move to a team with private repos
- On-call rotations or incident work
- A reorg that changed what someone owns

If you write "I noticed your commits stopped last week," you are guessing about someone's employment status from their activity log. If you guess right, it feels like surveillance. If you guess wrong, you have insulted someone who still has a job. Send every recent committer the same neutral, work-specific note whether or not you think they were affected.

## Which Yelp teams should you pitch, and why not Hatch?

Pitch platform, security, and data-infrastructure engineers. Skip Hatch and the AI teams, because Yelp's 8-K names AI-powered offerings as where it plans to focus, and people whose funding just went up have the least reason to leave.

Yelp completed its acquisition of Hatch, an AI-powered lead management platform, in February 2026. Hatch had about $25 million in ARR growing 70% as of November 2025. Yelp also said it "shifted our focus in Services lead management from Yelp Receptionist to supporting Hatch." That sentence tells you which work lost priority as well as which work gained it.

Infrastructure is the better target, but it is a small pool. In Refolk's index, only 39 profiles across the US, UK, and Germany combine an SRE, platform, infrastructure, or DevOps title with "Yelp." That is about 11% of the 342 software engineering profiles in those three countries. Site Reliability Engineer is the most common title (10 of 25), 9 of 25 currently work at Yelp, and they cluster in Mannheim, Los Angeles, and London.

That scarcity is the real problem for anyone hiring. Every platform team that read the 8-K is chasing the same few dozen people. Sourcing them by hand means cross-referencing commits against profiles one at a time, and you may still end up with candidates who are not open to moving.

When you do reach an infra engineer, lead with the work, not with Yelp's news:

- **Compute Infra / PaaSTA:** your orchestration stack, scale, and how much of it they would own.
- **Security / detect-secrets:** whether security is a funded function or one person handling tickets.
- **Data infra / Tron and mrjob:** your batch and pipeline problems, and whether those systems are a priority or legacy.

> **Site reliability engineering roles open now** - [see the open roles](https://www.refolk.ai/jobs/search?q=site+reliability&family=engineering)
>
> Platform and SRE engineers from Yelp's infrastructure teams will be comparing roles like these.

## How do you approach engineers who haven't job-hunted since 2020?

Make it cheap for them to say yes. Start with their public work, skip the résumé request, and offer a code conversation instead of a take-home. Yelp has not cut staff since 2020 and kept headcount approximately flat year over year in 2025, so many long-tenured engineers may not have run a job search in six years.

This is a hypothesis, not a measured fact, but the reasoning holds up. Someone who has spent six years at one company often doesn't have an updated résumé, hasn't practiced algorithm interviews, and has no recent read on what peers are paid. A standard multi-round loop with a take-home asks the most from the people with the least recent practice.

What tends to work better:

1. **Open with one specific commit or PR.** Show that you read their code and understand what problem it solved.
2. **Don't ask for a résumé in the first message.** Their GitHub and a short call are enough to qualify them.
3. **Replace the take-home with a code walkthrough.** Ask them to explain a PaaSTA or Tron change they shipped. It tests the same skills and respects work they have already done.
4. **Give compensation ranges early.** Someone who has not been on the market since 2020 needs a calibration point before they can take you seriously.
5. **Don't mention the layoff unless they do.** Your note should read the same to someone who was cut and someone who wasn't.

Every Yelp engineer with public commits will hear from a lot of recruiters this quarter. The message that gets a reply is the one that shows you understand their systems and won't make them reapply from scratch.

## FAQ

### How many people did Yelp lay off in November 2026?

Yelp's 8-K says the reduction is expected to affect "nearly 5%" of full-time employees. AIM Group reports about 250 people, a figure that comes from applying 5% to Yelp's 10-K headcount, but Yelp has not confirmed a final number. Restructuring charges are expected to be $14 million to $16 million, recorded mostly in Q4 2026 and Q1 2027.

### Were Yelp engineers affected by the layoff?

Yelp has not said which departments or locations were affected, so no public source confirms that engineers were cut. Approach Yelp engineers as people at a company that is reallocating budget toward AI, not as confirmed job seekers. That framing is more accurate, and it produces better outreach.

### Is it appropriate to contact engineers based on their GitHub commits?

Yes, if you mention the work and not their activity patterns. Referring to a specific PR they shipped is a reasonable opener. Guessing from a commit gap that someone lost their job is not, especially since PaaSTA shipped a release the day after the announcement. Send every recent contributor the same respectful message.

### Where are Yelp's engineers located?

In Refolk's index, US profiles with "Yelp" in the headline cluster in New York, San Francisco, and Bellevue, but only 2 of the top 25 currently work at Yelp. In Europe, they cluster in London and Berlin, and 18 of the top 25 are current staff. Infrastructure-titled profiles cluster in Mannheim, Los Angeles, and London.

---

_From the Refolk blog. More posts: https://www.refolk.ai/blog/index.md_
