Principle Engineer, 3D Reconstruction
Core
Lead the development of an offline 3D reconstruction system that builds high-quality 3D world geometry from vehicle sensor data in dynamic scenes.
Role type
Principal Engineer, 3D Reconstruction
Builds
Offline 3D geometric reconstructions and best-estimate ego motion from heterogeneous vehicle platforms and sensor suites
Domain
Autonomous vehicles, robotics, sensor fusion, mapping
Deliverable
production ML models | product features
Required skills
SLAM, 3D reconstruction, mapping, state estimation, lidar-based reconstruction, point-cloud registration, pose estimation, C++, Python, nonlinear optimization (pose graph optimization, factor graphs, bundle adjustment, ICP), sensor calibration, timing synchronization
Preferred skills
MS/PhD in Robotics/CS/State Estimation, GNSS/INS fusion, sensor-based odometry, modern machine learning for 3D reconstruction, foundation models, learned 3D representations, ROS, PCL, OpenCV, CUDA, ML data enrichment, autolabelling, validation-data generation
Technologies
C++, Python, ROS, PCL, OpenCV, CUDA
Responsibilities
Define technical architecture and roadmap for offline 3D reconstruction system; Lead development of MVPs to unblock high-priority internal use cases; Establish core technical interfaces, output representations, and tools; Partner with calibration, sensor fusion, perception, data, infrastructure, and validation teams; Set technical standards for reconstruction quality, failure analysis, uncertainty measurement, and dataset suitability; Mentor engineers contributing to the 3D reconstruction effort
Seniority
Principal, technical leadership & strategy
