Research Engineer / Scientist, Reinforcement Learning

Highlights

• Own the post-training recipe across SFT and RL for legal work.
• Improve long-horizon legal reasoning and agentic workflows.
• Design rewards and verifiers for challenging semi-verifiable legal tasks.

What they offer

• Offers Equity
• Offers Bonus

Ideal candidate

Hands-on RL researcher who owns training runs end-to-end and wants research to reach users fast. Thrives on ambitious, high-bar teams tackling long-horizon reasoning.

Requirements

• Hands-on experience post-training LLMs with SFT and RL including preference optimization
• Direct ownership of training runs and results with distributed training reliability
• Deep ML fundamentals and experience debugging divergence, reward hacking and instability
• Research experience in RL for language models or reward modeling