We are looking for an experienced Computer Vision & Edge AI Engineer to design and implement a real-time vision system running on NVIDIA Jetson (Orin, Xavier, Nano) devices. The project involves building a detection and volume estimation pipeline for industrial hoppers and exposing results through a lightweight API and dashboard. Responsibilities -Design and implement a computer vision pipeline: -Capture RTSP streams from IP cameras. -Detect & segment material inside hopper (YOLOv8/YOLOv10-seg). -Calibrate with OpenCV (homography, intrinsic/extrinsic parameters). -Calculate fill percentage and build an initial 3D volume prototype. -Optimize inference on Jetson devices using TensorRT / CUDA. -Build a backend service (FastAPI or Flask) to expose results as JSON/REST API. -Store data & snapshots in a lightweight database (PostgreSQL, InfluxDB, or TimescaleDB). -Implement trigger logic (e.g., stopped truck detection, passage through control zones). -Deliver clear documentation & support field testing. Requirements -Python (mandatory) with strong knowledge of OpenCV, PyTorch or TensorFlow. -Experience with YOLOv5/YOLOv8, instance segmentation, depth estimation (MiDaS, ZoeDepth), or stereo vision. -Proven hands-on work with NVIDIA Jetson devices (Nano, Xavier, Orin). -Strong background in edge optimization (TensorRT, CUDA) for real-time (15–30 FPS). -Backend/API development experience with FastAPI/Flask and SQL or time-series DBs. -Proficiency in Docker for packaging and deploying on the edge. To confirm you have read the full job description, please include the word JetsonReady in the first line of your proposal. If you have proven experience in Jetson-optimized computer vision pipelines, we’d love to hear from you.