Kārlis Dimza

I’m a Computer Vision / AI Engineer with several years of experience building production video analytics systems, with a strong focus on object detection, tracking, multi-camera tracking, ReID, pose estimation and edge AI deployment. I’ve worked on real-world applications across traffic analytics, sports, retail and surveillance, including multi-camera vehicle tracking, ANPR, fall detection, fisheye/overhead analytics and GPU-accelerated inference pipelines. My work covers the full lifecycle from dataset preparation and model training to evaluation, optimization and production deployment. I regularly work with Python, C++, Rust, PyTorch, Ultralytics YOLO, ONNX, TensorRT, NVIDIA DeepStream, OpenCV and GStreamer, including deployment on NVIDIA Jetson devices. I’ve built multi-camera tracking pipelines combining per-camera tracking, homography-based association and global identity matching, and I’ve optimized deep-learning models for real-time inference on constrained edge hardware. I’m particularly interested in senior Computer Vision / AI Engineer roles involving multi-camera tracking, sports analytics, intelligent video systems, edge AI or applied machine learning. I’m comfortable working independently in distributed teams and taking projects from research and prototyping through to production.