Azat K.

I build real-time computer vision, AR, and interactive 3D web applications focused on performance and practical deployment. **What I can help with:** • YOLO object detection — custom training and deployment • Real-time video processing — RTSP, IP cameras, CCTV, drones • AR / WebAR — image tracking, overlays, and interactive experiences • WebGL & Three.js — interactive 3D scenes, browser-based visualization, and AR experiences • Edge AI — NVIDIA Jetson Orin / Nano • Detection pipelines — people, vehicles, fire, and smoke **Experience includes:** • Training YOLO models on custom datasets • Building OpenCV-based tracking and video processing pipelines • Integrating computer vision with AR and interactive 3D applications • Developing browser-based 3D experiences using Three.js and WebGL • Working with MediaPipe and WebXR for AR and real-time interaction • Deploying ONNX models and working with GPU-accelerated inference • Saving detection events, images, and video clips **Tech stack:** Python, OpenCV, PyTorch, Ultralytics YOLO, ONNX, Three.js, WebGL, WebXR, WebAR, MediaPipe, JavaScript, React. I can help you build computer vision systems from scratch, integrate 3D and AR features into existing products, process live video streams, or create interactive browser-based 3D experiences. I focus on practical solutions, clean integration, and measurable results — not just demos. Ready to start quickly. I can begin with a small test project to demonstrate a working prototype on your data.