Lead ML Engineer - Lane & Route Network Mapping
Core
Architecting production-grade semantic and topological mapping stacks to enable autonomous vehicles to understand and navigate complex urban, suburban, and rural environments.
Role type
Lead ML Engineer (Autonomous Driving Mapping)
Builds
Vectorized HD maps, lane and route network graphs, and multi-camera BEV transformer models for real-time and offline mapping pipelines.
Domain
Autonomous Driving / Computer Vision / Robotics
Deliverable
production ML models
Required skills
Vectorized mapping networks (MapTR), BEV-based scene representation, Transformers, Graph Neural Networks (GNNs), Cross-modal calibration and fusion, Lane-level topology modeling, Computer Vision (detection, segmentation, tracking, depth estimation), PyTorch/TensorFlow, Python/C++
Preferred skills
Self-supervised/semi-supervised learning, Vision-Language Models (VLMs), Model optimization (quantization, pruning, distillation), Top-tier conference publications
Technologies
PyTorch, TensorFlow, MapTR, BEV, GNNs, Python, C++, Linux
Responsibilities
Lead research, design, and training of neural networks for vectorized mapping and multimodal fusion; Architect and implement production-grade mapping stacks; Drive feature development from inception to deployment; Own end-to-end data strategy including synthetic data and active learning; Develop evaluation frameworks for map accuracy and consistency; Collaborate with cross-functional teams to translate autonomy goals into system requirements.
Seniority
Senior, hands-on IC with leadership responsibilities
