The 72-Hour Playbook for Sourcing 700 Ex-Coinbase Engineers

On May 5, 2026, Brian Armstrong cut roughly 14% of Coinbase, about 700 people, and revoked system access the same day. By the time you read this, every recruiter on LinkedIn has already queued an InMail blast that opens with "saw the news." If you want to actually hire from this cohort, the next 72 hours are about doing the opposite of what everyone else is doing.

This is a concentrated pool of senior infra, payments, security, and management talent hitting the market in one window. It is also one of the worst cohorts in recent memory to compete on speed and generic outreach. Here is the playbook.

Why the standard layoff playbook fails this week

The reflex move after a public RIF is volume: pull the LinkedIn list, fire 300 InMails, hope 5% reply. That works when laid-off engineers are anxious and need income next month. This cohort is not that.

Coinbase's severance package gives affected US employees at least 16 weeks of base pay, plus two weeks per year of service, plus their next equity vest, plus six months of COBRA. Senior ICs are walking out with six to nine months of runway and intact health insurance. They are not desperate. They will read your message, register that it is a template, and archive it.

16+ weeks

The competition for their attention is also brutal. Layoffs.fyi reports roughly 100 tech companies have announced cuts this year totaling more than 92,000 employees. On the crypto side alone, the Coinbase cut joins Block (4,000 in February), Bolt (250 in April), Crypto.com (180 in March), Gemini (200 in February), and MARA Holdings (40 in April). Every fintech and crypto recruiter on Earth has the same target list open right now.

Speed is not your edge. Signal is.

What actually got cut, and why it matters

Read Armstrong's memo carefully. He is not just trimming for cost. He is restructuring around what he calls "AI-native pods," capping management at five layers below the CEO and COO, and requiring every leader to remain an individual contributor. In his framing, the company is killing "pure managers" in favor of "player-coaches," and experimenting with one-person teams directing agents that absorb engineer, designer, and PM responsibilities.

That framing has two consequences for sourcing.

First, the cohort is disproportionately weighted toward engineering managers and traditional EM/PM hybrids who did not fit the new pod structure. Most recruiters will reflexively chase the senior ICs and ignore the EMs. For a Series A founder who needs one person who can both ship and run a team of four, this is the highest-leverage hire on the market this week, and almost no one is going after it correctly.

Second, every remaining and laid-off Coinbase engineer has hands-on agentic-coding experience. Armstrong personally pushed GitHub Copilot and Cursor licenses to every engineer and demanded onboarding by the end of the week. Whatever you think of that mandate, it means this cohort is not "AI laggards." Mizuho's Dan Dolev told Bloomberg that the crypto winter is "probably the real reason for most of the cuts" and that AI is "likely an easy excuse." Either way, the engineers themselves are fluent in the tools that other companies are still piloting.

Pitch them accordingly. Do not apologize for asking them to use Cursor. Ask them what workflows they built around it.

Map the cohort in three lists, not one

Treat the 700 as three distinct sourcing problems.

1. The infra and payments builders

These are the people who worked on Base, USDC rails, Coinbase Prime, custody, the matching engine, and fraud/AML. Their skills are directly transferable to Stripe, Mercury, Ramp, Bridge (now part of Stripe), Anchorage, Robinhood, and the trust and safety teams at every major AI lab moving real money. The 2022 Coinbase cohort, when the company cut 18% of staff, dispersed exactly into those companies. Expect the same pattern.

The mistake is framing them as "crypto engineers." They spent years on regulated payments, high-throughput systems, and adversarial security. Pitching them on a stablecoin product when they would rather build fintech infra is how you lose the reply.

2. The displaced managers

The "pure manager" cohort. They will be invisible on most sourcing dashboards because their recent commits are sparse and their LinkedIn titles read "Engineering Manager" or "Senior EM." But they ran teams shipping regulated financial products. For early-stage founders, one of these hires can collapse three roles into one.

3. The about-to-found

A meaningful slice, plausibly 10 to 15 percent based on Coinbase's alumni-founder track record, will start companies instead of taking jobs. Coinbase itself has invested in 40 alumni-founded startups. The roster includes dYdX, Farcaster, Zora, and Goldfinch. Olaf Carlson-Wee, the first Coinbase employee, left in 2016 to start Polychain Capital, which went on to manage more than $300 million. Blake West was a senior engineer on Coinbase's core payments team and led engineering for the USDC savings product before founding Goldfinch. Fred Ehrsam co-founded Coinbase, then Paradigm, then Nudge.

If you are a founder reading this, your play with this slice is not a recruiting pitch. It is a co-founder intro or a seed check.

Where to actually find them in the next 72 hours

LinkedIn is the worst place to source this cohort right now, because it is where every other recruiter is. Three better surfaces:

GitHub commit history

Coinbase open-sources heavily: Base, kryptology, rosetta-sdk-go, build-onchain-apps. Pulling the contributor graphs on those repos surfaces the people who actually built the systems, not just the people whose titles said "Senior Engineer, Infrastructure." Cross-reference recent commit authors against current Coinbase employer tags, and you have a ranked list of builders, not headcount.

This is the kind of cross-surface query that is painful by hand. You want "engineers with 50+ commits to coinbase/base in the last 18 months who currently or recently listed Coinbase as employer, ranked by Solidity and Go depth, based outside the Bay Area." That is a plain-English question and it should return a plain-English answer, which is why we built Malinois: you describe the person and get a ranked shortlist across GitHub, LinkedIn, and the open web in one pass.

Farcaster and Warpcast

Disproportionately populated by ex-Coinbase engineers. Many of them are posting about what they are working on next in real time. A reply to a cast about a side project is worth ten InMails.

Layoffhedge and Layoffs.fyi

Both are tracking the Coinbase cut. The cohort itself congregates there to compare notes and severance details. It is also where they post when they are open to chat. Watching those threads for the next two weeks is a higher-yield activity than any sourcing tool dashboard.

A quick note on geography. Coinbase has been remote-first since 2022 with no physical headquarters. The cohort concentrates in New York, San Francisco, Toronto, and Bengaluru, but it is meaningfully distributed beyond those cities. If your sourcing strategy assumes Bay Area in-person coffee, you will miss most of the list.

The outreach itself: show up in week three

Here is the contrarian move. Do not send your first message this week.

Most recruiters will hit the cohort May 6 through May 10. Replies will be sparse because every engineer's inbox is a wall of identical sympathy notes. By May 20, the noise will have collapsed and the engineers who took two weeks off will be opening their laptops and triaging.

Show up then, with a thesis. Not "we are hiring senior backend engineers." Something like: "I read your 2024 talk on the rosetta-sdk-go integration. We are rebuilding payment reconciliation at [company] and the throughput problem you described is the one we are stuck on. Twenty minutes to compare notes?"

Speed is not your edge this week. The recruiter who shows up in week three with a thesis beats the one who shows up in week one with a template.

That message requires you to actually know what the engineer built. Pulling that kind of context across GitHub talks, blog posts, conference videos, and prior commit history for 200 candidates is the part that breaks most sourcing operations. It is also the part Malinois is designed to compress: ask for the cohort in plain English, get the per-candidate context inline, write the email yourself.

What to pitch them on

Three framings that work, ranked:

1. Payments and money movement at scale. This is the most direct skills transfer and the largest addressable opportunity. Stripe, Mercury, Ramp, Bridge, Modern Treasury, Increase, every neobank.
2. Trust and safety at AI labs. Custody, fraud, and AML translate cleanly to abuse, account takeover, and policy enforcement at OpenAI, Anthropic, and the major model providers. The compensation is competitive and the engineering problems are adjacent.
3. High-throughput infra at any company moving real assets. Trading platforms, prediction markets, marketplace giants. Coinbase's matching engine engineers have rare scale experience.

Three framings that lose:

1. "Come build the future of crypto." They just lived through revenue falling 21.6% in Q4 2025 and a $667 million quarterly loss while Bitcoin dropped more than a third from its $126,000 October peak. They know.
2. "We move fast and use AI." Armstrong already mandated Cursor and Copilot for everyone. You are not differentiated.
3. "We have great equity." Their next vest just landed in their severance. Cash and scope matter more than paper for the next six months.

The one-week timeline

• Day 1 to 2: Build the three lists (builders, EMs, likely founders). Do not message anyone. Watch Farcaster and Layoffhedge.
• Day 3 to 5: Per-candidate research. Pull GitHub history, conference talks, blog posts, prior employers. For sourcing ex-Coinbase engineers at any volume, this is where most teams stall and where a tool like Malinois pays for itself.
• Day 6 to 14: Founders and warm intros only. If you have a mutual connection, ask for the intro now, not later.
• Day 15 to 30: Cold outreach with thesis-driven first messages. Personalized to the specific system the engineer built.
• Day 30+: Founder cohort. Co-founder matching, seed conversations, advisor offers.

The 700 are not going anywhere fast. They have runway. The question is whether, when they are ready to talk, you are the one they remember as having actually understood what they built.

FAQ

How long is the hiring window for this Coinbase cohort?

Realistically, three to six months. With 16+ weeks of base pay, equity vest, and six months of COBRA, most of the cohort can comfortably wait until late summer or fall to make a decision. The senior ICs and EMs will be the slowest to commit because they have the most optionality. Plan for a long courtship, not a closing sprint.

Should I bother sourcing the engineering managers, or focus on ICs?

Focus on the EMs if you are a Series A or B founder who needs one person to ship code and run a small team. Armstrong is explicitly cutting "pure managers" in favor of player-coaches, which means the EMs on this list are by definition people who could not, or did not want to, stay hands-on at Coinbase's pace. Filter for the ones who still have recent commits in their GitHub history. Those are the player-coaches Coinbase let go of by accident.

Is GitHub really better than LinkedIn for sourcing ex-Coinbase talent?

This week, yes. Every recruiter on the planet is filtering LinkedIn for "Coinbase" plus "open to work." Almost no one is reading commit history on Base or kryptology to find the engineers who actually built the systems. The signal-to-noise ratio on GitHub for this specific cohort is roughly an order of magnitude better, and the engineers themselves notice when you reference their commits instead of their job title.

What about the engineers who want to start companies, not get hired?

Treat them as a separate funnel. If you are a founder, offer co-founder intros, advisor equity, or a seed check rather than a job. If you are a recruiter, the highest-value thing you can do is build a relationship now and place them at their second company in two to three years. The Coinbase alumni-founder track record (Polychain, dYdX, Farcaster, Zora, Goldfinch, Paradigm) suggests this slice is small in headcount but disproportionately important in long-term network value.