Anthropic
Research Engineer, Code RL (Reinforcement Learning)
$500K–$850K
San Francisco, CA +2 · Full-time · Posted Jun 11, 2026
- Location
- San Francisco, CA
New York City, NY
San Francisco, California, United States - Experience
- Years of experience required will correlate with the internal job level requirements for the position.
- Education
- Bachelor’s degree or an equivalent combination of education, training, and/or experience; relevant field of study as demonstrated through coursework, training, or professional experience.
- Visa sponsorship
- Visa sponsorship is available, but not guaranteed for every role or candidate; the company will make every reasonable effort to secure a visa for an offer recipient.
About the job
Advance coding capabilities in AI models by designing reinforcement-learning environments, rewards, and verifiers, running training experiments, and improving research pipelines.
What you'll do
- Design coding tasks, RL environments, reward signals, and verifiers.
- Run and interpret training experiments on frontier models to improve software-engineering capabilities.
- Improve the speed and reliability of training and experimentation pipelines.
What they ask for
- Strong software-engineering skills and deep Python expertise, including async/concurrent programming.
- Ability to own systems end to end and debug across the stack.
- Ability to shape experimental design, interpret results, and balance research with implementation.
- Care about code quality, testing, and performance; commitment to developing safe and beneficial AI systems.
Nice to have
- Experience with reinforcement learning, RLHF, post-training, or LLM finetuning; coding agents, sandboxes, evaluation harnesses, verifiers, or developer tooling.
- Background in program analysis, testing, verification, compilers, or formal methods; experience with PyTorch, distributed training, accelerators, or sandboxed execution.
Pay
$500,000 — $850,000 USD (base salary)
Tools
Python, PyTorch, CUDA, GPU, TPU
AI focus
Reinforcement learning to improve AI models’ ability to write, edit, test, debug, and ship software correctly, quickly, and safely.
Model-written summary · the original posting is authoritative