Founding Engineer - Software (Data/ML Infra)

We are looking for Founding Software (Data/ML Infra) Engineers to build the systems that power our model development from raw data collection through training infrastructure. You will design and maintain the pipelines that move, process, and serve terabytes of multimodal robotics data, and build the tooling and infrastructure that enable quickly shipping frontier embodied AI. You will work closely with the founders to shape both the technical and company roadmaps.

What You'll Do

• Build and maintain data pipelines for ingesting, processing, organizing, and querying large-scale multimodal robotics data, from teleoperation through training to deployment.
• Manage data versioning, lineage, and reproducibility across training runs.
• Build internal tools that let the team explore field logs efficiently.
• Set up and optimize training workflows - job scheduling, experiment tracking, resource management
• Optimize data loading and I/O performance to keep GPU utilization high during training.
• Build and maintain distributed training infrastructure as models and data volume grow.

Experience we look for

• Experience building and operating data pipelines for multi-modal data at scale
• Familiarity with cloud storage architectures and/or compute orchestration
• Experience profiling and optimizing I/O throughput and data loading performance
• Strong system design fundamentals with experience owning systems end-to-end
• Comfortable with common data/ML tooling

Nice to have

• Familiarity with robotics data formats and sensor data
• Familiarity with ML training workflows and frameworks
• Experience with data versioning and lineage tools (like DVC)