Senior Machine Learning Engineer - Large Driving Model (Behavioral Alignment)

We are seeking a Senior Machine Learning Engineer to drive behavioral optimization and alignment for our Large Driving Model (LDM). In this role, you will play a foundational part in the LDM Behavioral Alignment team, taking a holistic, vertical view of complex autonomous driving behaviors. Your objective is to analyze, define, and systematically improve vehicle actions by working fluidly across data, training, and model stacks. Rather than focusing on an isolated layer, you will target the root causes of insufficient driving performance and implement end-to-end solutions-ranging from advanced dataset curation to reinforcement learning (RL)-to directly deliver safer, smoother, and more predictable on-road behavior.

• Diagnose performance bottlenecks for targeted driving behaviors, establishing clear solution roadmaps across data, modeling, training recipe, and reinforcement learning loops.
• Design data mining strategies and metrics to understand targeted model behavior to drive performance insights, measure progress, and inform model fine-tuning and alignment.
• Design, experiment with, and deliver RL algorithms, Supervised Fine-Tuning (SFT) strategies, and RL reward structures to optimize on-road driving policies.
• Partner across the Autonomy organization with Infrastructure, Perception, and Motion Planning teams to implement high-impact solutions while successfully handing off long-term platform productionization to system owners.
• Provide technical guidance, code reviews, and mentorship for junior and mid-level engineers, fostering high standards for system design.

• B.S. or M.S. in Computer Science, Machine Learning, Robotics, or related fields.
• 5+ years of experience as an ML Engineer, Motion Planning Engineer, or Data Engineer within the LLM/VLM, autonomous driving, robotics, or computer vision space.
• Advanced proficiency in Python and deep learning frameworks (e.g., PyTorch), with a track record of training and fine-tuning large-scale models.
• Experience with Reinforcement Learning (RL), reward design, and policy optimization.
• Strong understanding of traditional planning concepts, edge cases, kinematic feasibility, and vehicle physics.
• Demonstrated ability to navigate highly ambiguous projects, shifting focus dynamically to solve vertical problems and deliver validated on-road impact.
• Strong cross-functional communication skills, with an ability to clearly balance technical trade-offs in a fast-paced environment.