LiDAR Characterization & Perception Data Engineer - Mountain View, CA

What you will be doing

• Sensor Characterization & FAE Support: Act as a technical liaison for customers. Assist with hardware integration and characterize 4D LiDAR performance (range, velocity, point cloud quality) in real-world environments to ensure optimal sensor deployment.

• Data Infrastructure: Develop Python and C++ infrastructure to ingest, validate, and organize data from collection vehicles and customer fleets: handling file verification, metadata extraction, and database writes.

• Calibration & Field Debugging: Run camera, LiDAR, and GPS calibration workflows on internal data collection vehicles and customer platforms. Debug hardware/software integration issues including sensor connectivity, logging failures, and network dropouts.

• Pipelines & Labeling: Build and maintain ETL pipelines that move high-volume sensor data to internal storage. Collaborate with external annotation vendors to package data for delivery, validate returned labels, and flag quality issues for the perception team.

• Operations Execution: Collaborate with operations and customer engineering teams to plan, support, and execute comprehensive data collection runs.

What You Have

• Bachelor's degree in Computer Science, Electrical Engineering, Optics, or a related field; or equivalent practical experience.

• 3+ years of industry experience blending software development (production Python and C++) with hardware/systems engineering (2+ years with a Master's).

• Hands-on experience with LiDAR, radar, or camera systems, including hardware troubleshooting, calibration, and point cloud analysis.

• Strong fundamentals in software engineering (data structures, system design) and experience designing or operating data pipelines at scale.

• Excellent communication skills, with the ability to translate complex technical issues between external customers, hardware engineers, and perception software teams.

• Familiarity with robotics or autonomous systems data such as ROS, sensor logs, binary formats, or perception datasets is a strong plus.

• Ability to work independently and cross-functionally in a fast-moving environment with ambiguous requirements.