OrthoScan is an AI-powered healthcare platform developed as a Final Year Project at Indus University that automates the detection of fractures and abnormalities from X-ray scans using deep learning. The system combines FastAPI, Celery, Redis, and Traefik in a fully containerized environment built with Docker and Docker Compose, providing scalability, asynchronous task management, and efficient load balancing. The AI model, originally trained in Keras, is converted to ONNX for lightweight, high-speed inference, enabling real-time diagnostic support even on limited hardware. Designed to connect patients, doctors, and hospitals through a unified interface, OrthoScan ensures faster, more accurate, and accessible radiology services while serving as a step forward in intelligent, technology-driven healthcare solutions.