Senior Machine Learning Engineer, Multimodal Perception

The Special Vehicle Compliance team develops the multi-modal perception, semantic reasoning, and driving intelligence that enables the autonomous vehicle to safely interact with high-stakes road actors. We are actively advancing our systems toward data-driven learned policies and end-to-end architectures, powered by large-scale closed-loop data engines.

Role overview: Autonomy MLE role focused on bridging perception and planning to develop and release robust learned driving policies. The primary objective of this role is to train, evaluate, and transition production-ready decision-making models into real-world autonomous navigation systems.

In this hybrid role, you will report to the Technical Lead Manager of the Special Vehicle Compliance team.

You will:

• Develop, evaluate, and release learned driving policies for complex navigation and yielding scenarios.
• Work cross-functionally at the intersection of perception and planning, translating multi-modal perception outputs into robust behavioral actions.
• Deploy learned models into closed-loop simulation environments, benchmark against strict safety metrics, and drive the transition of these models into production releases.
• Advance the transition from rule-based heuristics to scalable, data-driven learned policies.

• 2–5+ years experience training and releasing ML models in autonomous driving, robotics, or complex spatial AI.
• Hands-on experience working across both perception and planning stacks. Proficiency in learned driving policies (RL / imitation learning), PyTorch / JAX, closed-loop simulator evaluation, and a track record of releasing models to production.
• Familiarity with Vision-Language-Action (VLA) models and World Models is a strong plus
• Experience using foundation models and AI tools for scenario generation, evaluation analysis, and rapid experimentation.