Physical Intelligence
ML Infra Engineer, Data Systems
San Francisco · On-site · Full-time · Posted Sep 1, 2026
About the job
Build and operate data infrastructure that processes, stores, and delivers large-scale robot-learning datasets to training and evaluation systems.
What you'll do
- Design high-throughput data pipelines and batch or streaming workflows for multimodal datasets.
- Build storage, metadata, lifecycle-management, and data-movement systems; optimize data access and training-time throughput.
- Implement observability, validation, and safeguards, and collaborate with researchers, engineers, and roboticists.
What they ask for
- Strong software engineering fundamentals.
- Experience building distributed systems or large-scale data pipelines.
- Comfort with performance, memory, I/O, and storage efficiency.
- Familiarity with batch and/or streaming processing systems, object storage, and data format tradeoffs.
Nice to have
- Experience with large ML training pipelines or dataloading systems; knowledge of columnar or custom data formats.
- Experience with ClickHouse, Ray, Flink, Spark, or similar, and/or operating petabyte-scale datasets.
Tools
ClickHouse, Ray, Flink, Spark
AI focus
Build data infrastructure for large-scale robot learning, supporting model training and evaluation.
Model-written summary · the original posting is authoritative