The opportunity
We’re a team applying ML to push autonomy performance forward. As a Perception ML Engineer, you’ll develop and improve perception models, with ownership of modeling, training, and evaluation.
What you'll do
Improve scene understanding from real sensor data, including 3D and spatial representations
Generalize across sensor platforms, geographies, and rare driving scenarios
Meet requirements for model quality and onboard inference runtime
Applied ML Depth: Own model architecture, loss design, training, and experiments that deliver measurable perception improvements.
Collaborative Delivery: Partner with ML Data Infrastructure and autonomy engineers, contributing across the stack when needed while keeping primary ownership in modeling.
Develop and improve perception models, making choices about architecture,: representations, objectives, and training.
What they're looking for
- Design controlled experiments and ablations to understand model behavior and validate improvements.
- Diagnose model failures and distinguish limitations in modeling, supervision, training data, and optimization.
- Collaborate with autonomy and infrastructure engineers to turn sensor data: and failure analysis into targeted model improvements.
- BS in Computer Science, Robotics, Statistics, Physics, Math or another quantitative area.
