Skip to content
AI.info

jobs

Agent Post-Training Research

OpenAI’s Agent Post-Training team develops and improves the models behind agents in Codex, ChatGPT, the API, and other products. The work spans coding, tool use, computer use, multi-agent…

Company
OpenAI
Location
San Francisco, United States; London, United Kingdom
Status
Closed
Posted
2026-11-13 03:33 UTC

OpenAI’s Agent Post-Training team improves models used by agents in Codex, ChatGPT, the API and other products. Work spans coding, tool use, computer use, multi-agent coordination, long-horizon tasks, factuality, instruction following and calibrated reasoning. The person may lead a research direction, improve large training-run infrastructure, build evaluations that expose failures, or take a capability from experiments to launch. The posting asks for machine-learning, software, systems or statistics foundations plus practical experience with LLMs, reinforcement learning, post-training, evaluations, synthetic data or production ML. San Francisco or London, subject to approval.

Original job posting