OpenAI
Research Engineer, Codex
$380K–$500K
San Francisco · Full-time · Posted Aug 5, 2025
About the job
Improve the capabilities, reliability, and product fit of OpenAI’s agentic models through research, engineering, evaluation, training, and product integration.
What you'll do
- Design and run experiments to improve agent behavior, and build evaluations and environments that reveal failures.
- Develop and improve post-training systems, including data pipelines, graders, reward signals, and diagnostics.
- Partner across research, product, infrastructure, and safety teams to integrate improvements into model runs and shipped products.
What they ask for
- Strong technical fundamentals in machine learning, software engineering, systems, statistics, or a related field.
- Hands-on experience with one or more areas such as LLMs, RL, post-training, evaluations, model training, coding agents, or production ML systems.
- Ability to turn ambiguous behavioral problems into concrete experiments and analyze results.
- Clear communication and ability to work across research, product, infrastructure, data, evaluation, and safety teams.
Pay
$380K – $500K USD base salary
Tools
LLMs, RL, RLHF/RLAIF
AI focus
Train and improve agentic models, including work on coding, tool and computer use, multi-agent coordination, long-horizon tasks, factuality, instruction following, and calibrated reasoning.
Model-written summary · the original posting is authoritative