Research Infrastructure Engineer, Training Systems

📋 Description

• • Research Infrastructure Engineer, Training Systems at OpenAI, Inc. in San Francisco focuses on building and maintaining the systems layer that turns novel machine learning research ideas into runnable, measurable training workloads for large models, directly enabling model releases and research progress.
• • Day-to-day responsibilities include building and maintaining infrastructure for large-scale model training and experimentation, designing APIs and interfaces to simplify complex training workflows, improving reliability, debuggability, and performance across training and data pipelines, debugging issues across Python, PyTorch, distributed systems, GPUs, networking, and storage, and writing tests, benchmarks, and diagnostics to catch meaningful regressions.
• • The team works on research and systems that advance frontier models, often going beyond standard training recipes by building infrastructure to make new training approaches practical at scale, where systems work is directly tied to research progress through better tools, abstractions, and runtimes.
• • In this role, you can learn and achieve impact by enabling new model training approaches through systems innovation, deepening expertise in ML infrastructure and distributed systems, contributing to reliable and performant training pipelines, and working at the intersection of research and production engineering on cutting-edge AI systems.