Senior Runtime Engineer
Y0 is a runtime that executes other people's intentions against their real data. When it is slow, the product is bad; when it is wrong, the company is dead. You will own the execution layer — scheduling, state, replay, failure — and you will be one of the five people whose code every single run passes through.
What you will do
Own run execution end to end: scheduling, step state, retries, and fail-closed semantics when permissions are revoked mid-run.
Make replay real — any run, reproduced from its trace, byte for byte, months later.
Drive p95 run-start latency under two seconds and keep it there as load grows 10x.
Design the concurrency and isolation model so one tenant's runaway loop cannot touch anyone else's runs.
Carry a pager for the thing you build. We all do; the runtime is the company.
What we need
6+ years building distributed systems that other teams depended on in production.
Deep fluency in at least one of: Go, Rust, or the kind of TypeScript that knows what it costs.
You have debugged a distributed failure at 3am and can tell the story with the actual root cause, not the war-story version.
Strong opinions about state machines, idempotency, and exactly-once being a lie — and the scars to back them.
You write. Design docs, postmortems, commit messages people thank you for.
Nice to have
You have built a workflow engine, job scheduler, or durable execution system before — even a small one.
Experience metering and billing compute honestly.
You have run inference infrastructure and know where the latency hides.
Apply — we reply to everyone