Research Member of Technical Staff- Data Infrastructure

What You'll Do

• Arcitect, build, and scale a high-throughput data infrastructure that processes and manages billions of video clips with strong guarantees around reliability, latency, and cost efficiency
• Design and optimize large-scale storage systems (cloud object storage, databases, metadata stores) for multimodal datasets
• Build efficient indexing and retrieval systems to support fast dataset querying, filtering, and iteration for research and production use cases
• Develop observability frameworks for data pipelines including monitoring, alerting, failure recovery, and performance optimization
• Implement intelligent workload balancing and throughput optimization across distributed compute and storage systems
• Manage data artifacts, versioning, and lineage to ensure reproducibility and traceability across training runs
• Build internal interfaces and lightweight tools that enable researchers and engineers to explore, query, and analyze large datasets at scale
• Support integration and scalable deployment of vision-language models (VLMs) within data pipelines for screening, enrichment, or metadata generation

What We're Looking For

• 5+ years of experience in data infrastructure, distributed systems, ML infrastructure, or a closely related field
• Strong experience building and operating large-scale data pipelines (1B+ samples or petabyte-scale systems preferred)
• Deep understanding of distributed systems, databases, indexing strategies, and cloud storage architectures
• Experience optimizing data throughput, workload balancing, and cost-performance tradeoffs in cloud environments
• Strong skills in observability, monitoring, and production reliability for high-scale systems
• Software engineering fundamentals with the ability to own systems end-to-end, from design to production

Nice To Have (But Not Required)

• Experience managing large multimodal datasets
• Familiarity with ML training workflows and data lifecycle management
• Experience with robotics data formats or real-world sensor data (video, proprioception, teleoperation logs)
• Experience with data warehouse technologies (e.g., Snowflake, BigQuery, or Redshift) for large-scale data storage, querying, and analytics
• Familiarity with data versioning and lineage tooling (e.g., DVC, Delta Lake, or similar)