(1) Vendor needs to provide multi–camera person re–identification system.– A secure, real–time person re–identification (Re–ID) system to support workflow analytics.– This initiative focuses on identifying and tracking staff across multiple cameras in clinical environments while maintaining strict privacy compliance. – The system will incorporate fine–tuned deep learning models, encrypted embeddings, and edge–based deployment to ensure accurate and privacy–preserving identity management in hospital suites.– Build a real–time multi–camera ingestion and inference pipeline for person detection and tracking. – Fine–tune OSNet or similar deep learning pipeline on the client’s hospital–specific dataset to improve performance under clinical conditions. – Implement a multi–camera matching system to maintain consistent identities across fisheye and bullet cameras. – Ensure full privacy compliance, including no storage of original frames and encryption of all embeddings. – Deploy the Re–ID system on hospital–owned edge devices (e.g., NVIDIA AGX Orin) for low–latency inference. – Provide documentation and APIs for integration into workflow analytics tools. (2) All the questions must be submitted no later than December 22, 2025.