About Ultra

Overview

What You'll Do

• Own our data platform and scale it to ingest large amounts of video streams and make them available for training in real time
• Own our ML infrastructure stack end-to-end, including distributed training, experiment tracking, and cluster management
• Build out data management systems and high-performance data access layers on top of PB+ scale multimodal data
• Collaborate with the rest of our research team on designing and running experiments as we rapidly improve our policies’ capabilities
• Maintain high availability and reliability of critical infrastructure

Who You Are

• You have deep experience building and operating large-scale data systems (PB+ scale), ideally in domains like autonomous vehicles or robotics
• You're comfortable with real-time processing, streaming, and event-driven systems
• You have built deep learning training and inference systems at scale
• You thrive in a high-trust, high-autonomy environment. You don't need to be micromanaged on what the top priorities are at any given moment
• You're hungry for impact and personal growth, and like to have fun in the pursuit
• Deeply passionate about robotics and physical AI

Bonus Points

• Experience working on large-scale autonomous vehicles datasets or similar robotics domains
• Hands-on experience with distributed ML training at scale (1000+ GPU hours)
• Familiarity with video codecs, compression, and efficient video storage/retrieval systems
• Experience with reinforcement learning, imitation learning, or VLA model training pipelines
• Experience working with hardware systems in a production environment