Monitor-AI

Monitor-AI is a local-first AI-powered CCTV intelligence platform designed to transform compatible existing DVR/NVR camera systems into intelligent workplace monitoring systems. It combines YOLO11n for person detection, YOLO11n-Pose for 17-keypoint human posture estimation, and YuNet with SFace for face detection and recognition. The system classifies sitting, standing and unknown postures, tracks activity durations, and stores records locally using SQLite. Monitor-AI also integrates Meta's Model API, bringing Muse Spark, Muse Wonder and SAM 3.1 into a unified application for AI assistance, advanced reasoning and visual segmentation. Its core computer vision pipeline runs locally on the user's hardware. CCTV streams and biometric data are not sent to Meta. The application is designed to make AI-powered workplace monitoring more accessible through configurable camera integration and CPU-based processing.