Software Engineer, ML Infrastructure

What you'll do

• Build and operate core model training and dataset management infrastructure for perception and autonomy use-cases.
• Develop systems for indexing multi-modal sensor data and orchestrating training across cloud and on-premise compute.
• Maintain and improve internal perception platforms for dataset querying, discovery, and model evaluation.
• Develop and iterate on training libraries and SDKs to support perception engineers across the company.
• Integrate frontier models into data curation pipelines for augmentation, annotation, and synthetic data generation.
• Build systems for real-time monitoring of training jobs, model registries, and full ML lifecycle traceability.
• Partner with business line engineering teams to deliver custom data solutions and infrastructure improvements.

What we're looking for

• Build and operate core model training and dataset management infrastructure.
• Improve perception engineer-facing tooling and SDKs.
• Integrate frontier models for data augmentation, curation, and synthetic data generation.
• Enable observability and traceability of the full ML lifecycle.
• Engage with customers to understand and deliver specific data requirements.
• Strong proficiency in modern Python and exposure to Rust/the Rust ↔ Python toolchain (preferred).
• Experience with modern deep learning frameworks and distributed training paradigms (preferred).
• Experience with MLOps tooling, Kubernetes, and U.S. Person status for export-controlled data (preferred).