About this role
Polymath is an applied research lab focused on advancing long-horizon agent capabilities through reinforcement learning. They are seeking Software Engineers to build simulation environments, tasks, and verifiers that challenge frontier models and improve autonomous agents.
Responsibilities:
- Build diverse, high-fidelity environments that test agents in realistic settings
- Design complex tasks that require long-horizon reasoning, tool use, and adaptation
- Develop robust verifiers that reliably measure agent performance
- Improve infrastructure and tooling to run, debug, and improve Polymath’s environment simulation platform
- Work closely with the research team to identify failure modes and turn them into new tasks and benchmarks
Requirements:
- Have strong engineering fundamentals
- Enjoy building from first principles and solving open-ended technical problems
- Have high agency and a strong bias toward shipping
- Have a high quality bar and care about building robust systems