Decentralized Remote Collaborative System with Federated AI (DRCSFA)

A secure, decentralized platform for remote collaboration, proctoring, and assistance. Features WebRTC video/screen sharing, Electron-based remote desktop control, local and federated anti-cheat monitoring, audited session chat, and immutable blockchain audit anchoring.

Architecture

• Frontend: Next.js (React), TailwindCSS, Radix UI Primitives
• Backend: Node.js, Express, Socket.IO
• Desktop Client: Electron, RobotJS (for native OS input)
• Database: MongoDB (with local JSON fallback)
• AI Engine: TensorFlow.js, MediaPipe Vision, federated client update aggregation
• Blockchain: Hardhat (Local Ethereum node for anchoring audit logs)

Prerequisites

• Node.js (v22 recommended)
• Docker & Docker Compose (for the full local stack)
• Windows OS (for the Electron Host App to execute robotjs inputs correctly)

Quick Start Setup

1. Install Dependencies

npm install

2. Configure Environment

Copy the unified example configuration to your local .env:

cp .env.example .env

Ensure you generate a strong JWT_SECRET (at least 32 characters) for production use.

3. Start the Full Local Stack

Use Docker Compose to start MongoDB, Hardhat, MinIO, the backend, and the frontend:

docker-compose up -d

The Web UI will be available at http://localhost:3000.

4. Manual Development Mode

If you do not use Docker, ensure MongoDB is running or let the backend fall back to mock storage. Then open two terminal windows:

Terminal 1 (Next.js Frontend):

npm run dev

Terminal 2 (Express + Socket.IO Backend):

npm run server

The Web UI will be available at http://localhost:3000.

5. Launch the Electron Desktop Client

For users who need to share their screen and grant remote control access, they must use the Electron app:

npm run electron:dev

Production Deployment

See DEPLOYMENT.md for the step-by-step runbook (secrets, regional TURN relays, Polygon audit contract, mobile builds).

1. Build the Next.js application: npm run build
2. Deploy the exported static frontend from out/ or use the existing Vercel configuration.
3. Deploy the backend separately and set NEXT_PUBLIC_BACKEND_URL to that backend URL.
4. Ensure the backend process has production JWT_SECRET, MongoDB, object storage, email, blockchain, and TURN settings.
5. HTTPS and a proper TURN server are strictly required for production WebRTC.

WebRTC Configuration

By default, the application uses public Google STUN servers. For production reliability across restrictive NATs/Firewalls, provide your own TURN server credentials (e.g., Twilio or Coturn) in the .env file.