Computer Vision Engineer — Quality Inspection

We build AI-powered quality inspection systems for industrial manufacturing. Currently deployed on 6 production lines across 3 suppliers in Germany and Poland — our models process camera feeds at 30fps and flag defects in under 80ms. Any slower and the line doesn't stop in time. We need a senior CV engineer to own model development for our Q3 expansion: two new lines (injection moulding and PCB assembly) with defect types we haven't handled before. Full ML cycle ownership: dataset creation from our annotated image archive, model training (YOLO-based and custom architectures), inference optimisation for NVIDIA Jetson AGX, and deployment alongside our hardware team. We already have annotation tooling, a CI pipeline for model versioning, and a benchmarking rig on-site. What we're missing is someone who can drive the model side independently. If you've shipped defect detection in production, tell us specifically what you built and what precision/recall numbers you hit.