Edge AI inference pipeline combining NVIDIA Jetson for on-device visual processing with a Django REST/WebSocket backend and Flutter mobile client. Camera feeds flow to the Jetson for real-time inference, results stream to Django which persists data in MinIO object storage and a metadata database, then serve insights to Flutter clients via REST and WebSocket APIs. This architecture minimizes latency by processing at the edge while maintaining centralized storage and cross-platform mobile access. Fork and customize this diagram on Diagrams.so to adapt inference models, storage backends, or mobile frameworks to your use case.