Agent Post-Training, API & Power Users
OpenAI · San Francisco
CHECKED 1H AGO
Frontier labonsite$380k–$500k
At a glance
Highlights
- Train frontier agents shipped in Codex, ChatGPT, and the API.
- Turn real developer workflow failures into shipped model improvements.
- Own agentic behavior projects from discovery through launch readiness.
Ideal candidate
High-agency generalist who thrives on ambiguous model-behavior problems and cares deeply about developer experience, turning qualitative failures into concrete shipped improvements in frontier agents.
Requirements
- Strong technical fundamentals in ML, software engineering, systems, statistics, or applied research
- Hands-on experience with LLMs, post-training, RL, evals, coding or tool-using agents, or production ML
- Strong taste for model behavior from transcripts, traces, and eval failures
- Comfort with ambiguous capability problems spanning data, training, evals, and product
- Deep care for developer and expert-user experience in real workflows
- Ability to work across research, product, infrastructure, data, evals, and safety
- Willingness to build load-bearing systems and processes when needed
- Motivation to train and ship models that make agents genuinely useful
Lab profile
About OpenAI
Builds safe frontier AI models and products like ChatGPT for broad benefit.
What they build
- ChatGPT
- OpenAI API
- Frontier AI models
Known for
- Frontier-scale distributed ML training
- Massive GPU fleet and data center operations
- Exabyte-scale data lakes and streaming
- AI alignment and safety research
Key facts
- Operates one of the world's largest GPU fleets