← All jobs
Research Engineer, Post-training
Anthropic
San Francisco, CAOn-siteFull-timeMid-level3+ years$260,000 - $340,000 + equity
Post-training sits between research and production — you'll build the infrastructure researchers use to actually run their ideas at scale.
Requirements
- 3+ years of ML engineering, with hands-on RLHF or fine-tuning experience
- Strong software engineering fundamentals, not just research scripts
- Comfortable working in a fast-moving, high-ambiguity research environment
Responsibilities
- Build and run training infrastructure for post-training experiments
- Design evaluation suites that catch regressions before external release
- Collaborate directly with researchers on the next model's training recipe
Skills
- Python
- PyTorch
- LLM fine-tuning
- Distributed systems
- CUDA