jobs

Training - Runtime Foundations Engineer

About the TeamTraining Runtime designs the core distributed runtime that powers everything from early research experiments to frontier-scale model runs. We build robust, scalable, high performance components to support our distributed train

In inglese

Company

OpenAI

Location

San Francisco

Status

Open

Posted

2026-03-09T18:29:12.507+00:00

OpenAI's Runtime Foundations team, inside Training Runtime, hires an engineer to build the distributed runtime that orchestrates machine-learning training across clusters of hundreds of thousands of machines. Work includes designing and maintaining orchestration and monitoring software, profiling the stack, and improving reliability, observability and fault tolerance, mostly in Rust with Python. The posting asks for distributed-systems experience, proficiency in Rust or another systems language such as C++, Linux debugging and profiling skills, and comfort with asynchronous, concurrent systems. Based in San Francisco, hybrid, three days a week in office.