Agent Post-Training, Computer Use Research

As a member of the Agent Post-Training team at OpenAI, you will teach frontier models to operate computers, navigate browsers, and complete complex long-horizon workflows with reliability. This role involves designing experiments, building evaluation environments, and managing the post-training stack to shape the capabilities shipped in products like ChatGPT and Codex. You will collaborate across research, engineering, and product teams to define what next-generation agents should achieve and ensure they are safe and useful.