Description
Thinking Machines is hiring a Research Scientist for its Coding Agents team to advance agentic coding capabilities through post-training research. The role involves designing and running reinforcement-learning training jobs, building sandboxed coding environments and reward signals, generating and curating synthetic coding data, developing evaluations, debugging large-scale RL runs, and collaborating with infrastructure and evaluation teams. The position is based in San Francisco, offers $350,000–$475,000 USD annually, and includes visa sponsorship and health, dental, and vision benefits.
