About the job
About the Foundations Retrieval Team
The Foundations Research group at OpenAI explores new approaches that could shape artificial intelligence for years to come. The team focuses on improving the science and data behind model training and scaling, especially for future advanced models. Areas of focus include data utilization, scaling laws, optimization strategies, model architectures, and efficiency improvements.
Within Foundations, the Search team builds agentic search solutions. This group works closely with others to design interfaces between models and the core search stack, serving, indexing, and retrieval, so model intent leads to reliable, real-world results. The team develops large-scale systems to transform and index massive information sources, enabling models to reason over global knowledge. Close collaboration with researchers helps move new modeling ideas into production quickly, changing how intelligent systems discover and synthesize information at scale.
Role Overview
OpenAI is hiring a Software Engineer with expertise in retrieval system development and scalability for its San Francisco office. This role involves working with researchers and engineers to build infrastructure that lets models access the right information when needed. Responsibilities include designing and operating indexing systems, retrieval pipelines, and serving layers.
Work in this role will directly improve retrieval capabilities across OpenAI’s research and products, with a strong influence on system performance, reliability, and scalability.
What You’ll Do
- Develop and scale retrieval infrastructure, including indexing, serving, and query execution.
- Build low-latency, high-throughput systems for real-time model interactions.
- Work with research teams to bring embedding and retrieval methods into production.
- Support dense, sparse, and hybrid retrieval pipelines.
- Maintain system performance, reliability, and observability at scale.
- Collaborate with Pretraining, Inference, and Product teams to deliver end-to-end retrieval solutions.
- Help develop model-system interfaces for agentic workflows.
Who We’re Looking For
- Experience building and scaling distributed systems.
- Background in developing high-performance, low-latency systems.
- Hands-on work with indexing and retrieval techniques.
- Familiarity with hybrid retrieval systems.
- Comfort working collaboratively across multiple teams.

