Own the serving stack for a frontier-scale inference platform handling 40B tokens a day.
What you'll do
- Ship production work in your first month — this team does not do six-week onboarding decks.
- Own a meaningful surface end-to-end, from design docs to on-call.
- Work directly with the founding team; write things down; argue in documents, not meetings.
What they're looking for
- Demonstrated depth in AI — shipped work beats credentials.
- Demonstrated depth in Python — shipped work beats credentials.
- Demonstrated depth in Inference — shipped work beats credentials.
- Comfort with ambiguity and a bias for small, fast releases.