ML Infra Engineer, Data Systems
Remote • San Francisco • FullTime
Posted 1mo ago
Job Location
San Francisco
Tech Stack
Remote Work Policy
Fully remote
Employment Type
FullTime
Categories
Machine Learning Engineer
About the job
Physical Intelligence is building general-purpose AI for the physical world, developing foundation models and learning algorithms for robots and future devices. As an ML Infra Engineer focused on Data Systems, you will construct and manage the data infrastructure essential for large-scale robot learning. Your work will bridge raw data sources and the training/evaluation phases, ensuring speed, accuracy, and reliability as systems scale. This role is a systems-focused position at the intersection of distributed systems, storage, and machine learning infrastructure.
Responsibilities
- Design and build high-throughput pipelines for validating, transforming, and featurizing raw multimodal data.
- Operate large-scale batch and streaming workflows over massive datasets.
- Design object storage layouts, metadata systems, and efficient access patterns, selecting appropriate file formats for performance and scalability.
- Build systems for data lifecycle management, including backfills, dataset rebuilds, garbage collection, and large-scale transformations.
- Optimize dataloaders, sharding, prefetching, caching, and throughput to reduce the time from data arrival to model training.
- Build scalable metadata stores for datasets, annotations, and training artifacts.
- Implement efficient data movement solutions across clusters and environments for petabytes of data.
- Develop observability, validation, and guardrails to prevent silent data regressions and ensure operational correctness.
- Collaborate with cross-functional teams of researchers, engineers, and roboticists to translate data needs into robust systems.
Requirements
- Strong software engineering fundamentals.
- Experience building distributed systems or large-scale data pipelines.
- Ability to reason about performance, memory, I/O, and storage efficiency.
- Familiarity with batch and/or streaming processing systems.
- Experience with object storage systems and data format trade-offs.
- An ownership mindset, with the ability to design, build, operate, and iterate on systems end-to-end.
- Enjoy working closely with researchers and unblocking fast-moving projects.