Agent Post-Training, API & Power Users

Responsibilities

• Design and run experiments to improve model behavior in API and power-user workflows, including function calling, tool use, and planning
• Build evals, graders, and environments from real developer workflows to turn failures into training data
• Partner with API and power-users to identify behavior gaps and convert product signals into post-training interventions
• Improve model behavior when composed into systems, ensuring reliability in tool use and multi-step tasks
• Own end-to-end model behavior projects from qualitative failure analysis through data generation and training experiments
• Develop feedback loops using power-user traces and API usage patterns to discover agentic model failures
• Debug hard failures in shipped models by analyzing traces, evals, and training data
• Work on early-training and alignment interventions, including data mixtures and synthetic data

Requirements

• Strong technical fundamentals in ML, software engineering, systems, or statistics
• Hands-on experience with LLMs, post-training, RL/RLHF/RLAIF, or production ML systems
• Experience with evals, graders, synthetic data, or coding/tool-using agents
• Ability to form concrete hypotheses from qualitative model failures or API interactions
• Comfortable working across research, product, infrastructure, and safety boundaries

About the Company

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products.