● Developed C++ and Python evaluation tooling that improved simulation accuracy and reduced debugging time for on-road autonomy disengagements.

● Built internal automation tools (Python/Bash + Buildkite) to streamline simulation regression testing, reducing weekly manual engineering effort and increasing test throughput.

● Integrated hardware-in-the-loop test flows with simulation stack to enable large-scale validation against real onboard vehicle runtime services.

● Led development of targeted regression datasets in collaboration with Perception teams, enabling scalable validation of ML-driven autonomy features(traffic lights, lidar artifacts, bounding box occlusions).

● Developed Python and Bash tooling to integrate sensor features (cameras, lidar, radar, IMU, and GPS) and automate system configuration validation, improving fleet reliability and uptime.

● Implemented low-latency C++ IMU driver integrated into real-time pub/sub architecture, improving localization responsiveness for safety-critical behaviors.

● Managed backend infrastructure and automation for Docker-deployed internal Flask calibration app used by hardware, operations, and perception teams.

Supported sensor system integration across autonomous vehicle platforms and built Arduino-based test

tooling to improve first-article inspection reliability and maximize vehicle operation efficiency.