Advance CCMS (Vigil)
October 13, 2026
AI

Advance CCMS (Vigil)

Cross-platform command center management system that brings video surveillance, access control, alarms, IoT automation, offline video analytics and a Claude-powered AI copilot into one security operations console.

React 19TypeScriptViteNode.jsFastifySQLiteWebSocketsWebRTCgo2rtcPythonOpenCVClaude APIReact Native (Expo)ElectronLeafletModbus TCPMQTTDocker

Overview

Vigil (formerly Advance CCMS) is a security operations console for command centers. It connects any brand of camera, NVR or DVR through ONVIF, native Hikvision, Dahua and Axis drivers or plain RTSP, and adds access control, alarms with SOPs, recording and evidence-grade playback, site maps, IoT barriers and sensors, and virtual guard patrols. A Python + OpenCV engine runs 13 types of video analytics offline with no licence fees, and Vigil AI — built on the Claude API with tool use — answers questions about devices, alarms and access logs, reads live camera frames and drafts shift reports, while every door or alarm action waits for an operator to confirm. It runs as a web app, a PWA, an Electron video-wall shell and a native React Native mobile app.

Key Achievements

  • Built a pluggable driver layer (ONVIF, Hikvision, Dahua, Axis, RTSP, HTTP relay) behind one interface for probe, snapshot, PTZ, door control and event subscription
  • Integrated Vigil AI using Claude tool use to query device health, alarms and access logs, inspect live camera frames and draft shift reports, with confirm-before-acting on every door or alarm action
  • Developed a free, offline video analytics engine in Python and OpenCV with 13 rule types — intrusion, tripwire, loitering, crowding, abandoned objects, wrong-way, tamper, flame and more
  • Delivered low-latency live video with WebRTC through a go2rtc media gateway, two-way intercom audio, PTZ presets and tours
  • Implemented recording and NVR playback on a 24-hour timeline with alarm markers, and MP4 clip export with SHA-256 hashes for evidence
  • Added access control with card, PIN, fingerprint and face credentials, and IoT automation for barriers, sensors and relays over Modbus TCP, MQTT and HTTP
  • Shipped nested site maps and floor plans with live device markers, virtual camera patrols, scheduled snapshot reports, and a React Native app with biometric unlock and alarm push notifications

Advance CCMS (Vigil)

Vigil is a cross-platform command center management system: a security operations console that brings video surveillance, access control, alarms and an AI assistant into one screen.

Command center dashboard
Command center dashboard

Highlights

  • Any brand of camera, NVR or DVR — ONVIF, native Hikvision / Dahua / Axis drivers, or plain RTSP.
  • Access control — doors, cardholders and credential rules with card, PIN, fingerprint and face; each door can have a linked camera for alarm video and evidence snapshots.
  • Real-time operations — live video wall with WebRTC and PTZ, two-way audio, an alarm queue with SOPs and operator notes, and an event journal with CSV export.
  • Recording and playback — a built-in recorder plus NVR playback on a 24-hour timeline with alarm markers; download any range as an MP4 with a SHA-256 hash for evidence.
  • Site maps and floor plans — nested campus → building → floor plans with live camera, door and device markers.
  • IoT, barriers and automation — sensors and relays over Modbus TCP, MQTT and HTTP, and when → and → then automation rules.
  • Virtual patrol — guard tours through the cameras with checkpoints, snapshots as proof and missed-round alarms.
  • Video wall
    Video wall
    Site maps
    Site maps

    Vigil AI

    An operator assistant built on the Claude API with tool use. Operators ask questions in plain language about device health, alarms and access logs. It can look at live camera frames, draft shift reports and propose door or alarm actions — but it never unlocks a door on its own. Every such action waits for an operator to press Confirm.

    Vigil AI
    Vigil AI

    Free Video Analytics

    A Python + OpenCV engine runs 13 rule types drawn directly on the camera image: intrusion, tripwire, counting, loitering, crowding, abandoned and removed objects, wrong-way, running, camera tamper, frozen or lost video, and flame. It runs on the site's own CPU, offline, with no licence fees.

    Video analytics
    Video analytics

    Architecture

  • Clients: React web app (also a PWA), Electron desktop shell for kiosk and video-wall windows, React Native mobile app
  • Server: Node.js + Fastify with auth and roles, devices, events and alarms, access control, layouts, audit and Vigil AI
  • Storage: SQLite (node:sqlite)
  • Video: go2rtc media gateway registers streams on demand and serves WebRTC to browsers
  • Analytics: Python + OpenCV engine pulls rules and pushes events with evidence JPEGs back to the server
  • Offline-ready: icons and fonts are bundled, with no CDN, so it runs on closed networks
  • Mobile app
    Mobile app

    Have a project in mind? Let's build it.

    I'm open to full-time roles and interesting collaborations in Singapore, India and remote.

    Sathishkumar Ranganathan

    Full-stack software engineer building enterprise platforms, mobile apps and AI-powered tools with Laravel, React, Node.js, Flutter and AWS.

    Connect

    Keyboard Shortcuts

    ⌘K
    Search
    T
    Toggle theme
    2026 Sathishkumar Ranganathan. All rights reserved.
    Built withusing Next.js & Tailwind