Building Hivemapper’s large-scale mapping pipeline to localize map features such as traffic lights and stop signs with <1m accuracy and extract insights from imagery and sensor data.
Developing optimized ML pipelines for low-power IoT dashcams, enabling real-time object detection on constrained hardware.
Building sensor fusion + Computer Vision models to detect traffic violations and incidents on device in real time.
Building a vehicle fleet management platform, tracking thousands of vehicle locations, statuses, trip history, and safety events in real time.
