OpenAI
Technical Lead Manager - Training Runtime, Data(set) Movement
$380K–$500K
San Francisco · Hybrid · Manager · Full-time · Posted Jun 8, 2026
About the job
Own and build the dataset-read platform for large-scale model training, with the role later expanding to broader in-cluster data movement systems.
What you'll do
- Design and build unified dataset reads, APIs, storage contracts, versioning, and migration paths across training frameworks.
- Develop reliability, caching, recovery, debugging, and visualization tools for training data pipelines; write and review production code.
- Coordinate with research and infrastructure teams, then lead broader data movement work including checkpoint loads/saves and snapshot transfers.
Nice to have
- Experience with stateful iterators, checkpoint/restart semantics, caching, remote services, or high-throughput storage reads.
- Rust or C++ experience; experience with multimodal, video, reinforcement learning, or pretraining data pipelines.
Pay
$380K – $500K USD base salary
Tools
Python, Rust, C++, torch.utils.data
AI focus
Builds infrastructure that supplies data for model training, including pretraining, reinforcement learning, and multimodal training.
Model-written summary · the original posting is authoritative