Principal Software Engineer, Agent Harness Bridge

ABOUT THE TEAM OpenAI's research training infrastructure powers how our frontier models are trained and evaluated. The Agent Harness Bridge team sits at the boundary between the Codex harness, research infrastructure, and internal agent systems, ensuring the same agentic coding runtime used in product can also support large-scale training workloads. This team owns the integration layer that connects harness capabilities into the training stack. The work is highly cross-functional and high leverage: researchers depend on it to run reliably, platform teams depend on it to evolve cleanly, and failures in this surface can materially affect training velocity and correctness. ABOUT THE ROLE We're looking for a Principal Software Engineer to lead the architecture and evolution of the Agent Harness Bridge. You'll own a critical interface between research and engineering, building the systems, APIs, and operational patterns that let researchers use agentic coding infrastructure safely