I build computer vision and OCR systems that turn images and documents into clean data. Work includes OCR extraction at 97 to 99% accuracy, container ID reading with ISO 6346 validation, real-time glove compliance checks, and item detection with YOLOv8 and YOLOv11. Stack: Python, OpenCV, Roboflow, PaddleOCR and GPT-4 Vision. You get a trained model, a pipeline and a simple way to run it.