Distributed Training Engineer, Sora

OpenAI is seeking a Distributed Training Engineer for their Sora team, focusing on making video a key capability of OpenAI's foundation models. This role combines research and product development, working to improve training throughput for internal frameworks and enable researchers to experiment with new ideas. The position requires expertise in distributed systems, ML optimization, and strong engineering skills.

The ideal candidate will work on optimizing state-of-the-art AI models, writing robust machine learning code, and maximizing supercomputer performance. This is a hybrid position based in San Francisco, requiring 3 days in the office per week, with relocation assistance available.

As part of the Sora team, you'll be at the forefront of video model development, working to understand and expand model capabilities while ensuring reliability and safety. The role involves close collaboration with researchers, implementing cutting-edge techniques, and maintaining high performance standards.

OpenAI offers competitive compensation ($380K – $555K plus equity) and the opportunity to work on transformative AI technology. The company is committed to diversity, equal opportunity, and creating AI that benefits all of humanity. Join a team that's pushing the boundaries of AI while focusing on safety and human needs at its core.

The position requires expertise in multi-modal ML pipelines, distributed systems, Python programming, and training optimization. You'll be working with a team dedicated to understanding and improving training dynamics while maintaining code quality and system performance.