Careers
We build auto‑improving inference endpoints from the lens of data research. All roles are on‑site in San Francisco or New York.
-
ML Systems
- Post‑trained open‑weight models of 8B+ parameters
- Inference optimization and kernel development
- Led research that touched the model end to end
- Experience in resource‑constrained environments
-
Applied AI
- Post‑trained open‑weight models of 8B+ parameters
- Opinions on post‑training frameworks (SkyRL, Miles) and compute providers
- Led research that touched the model end to end
- Experience in resource‑constrained environments
- Bonus: online RL post‑training
-
Core Engineering
- Run long‑horizon agent workflows processing 1T+ tokens a week
- Distributed systems, data pipelines, performance optimization
For the ML roles, we want hands‑on model work like post‑training, kernels or transfer learning. Apps built on API calls don’t count.