About the job
As a member of the Codex team, you will advance the capabilities, performance, and reliability of AI coding models through a combination of research, experimentation, and system optimization. You’ll collaborate with world-class researchers and engineers to develop and deploy systems that help millions of users write better code, faster—while also ensuring these systems are efficient, cost-effective, and production-ready.
Responsibilities
Design and run experiments to improve code generation, reasoning, and agentic behavior in Codex models.
Develop research insights into model training, alignment, and evaluation.
Hunt down and address inefficiencies across the Codex system stack—from agent behavior to LLM inference to container orchestration—and land high-leverage performance improvements.
Build tooling to measure, profile, and optimize system performance at scale.
Work across the stack to prototype new capabilities, debug complex issues, and ship improvements to production.
Qualifications
Minimum
No minimum qualifications listed.
Preferred
Are excited to explore and push the boundaries of large language models, especially in the domain of software reasoning and code generation.
Have strong software engineering skills and enjoy quickly turning ideas into working prototypes.
Think holistically about performance, balancing speed, cost, and user experience.
Bring creativity and rigor to open-ended research problems and thrive in highly iterative, ambiguous environments.
Have experience operating across both ML systems and cloud infrastructure.