Voice Authenticator

Prototype Zoom App for real-time speaker authenticity checks during meetings. The project combines a Node/Express Zoom App shell with a Python backend that streams voice-verification status updates to the client.

Features

• Zoom App integration using the Zoom Apps SDK.
• OAuth-backed Node/Express server.
• Python Flask service for realtime verification status.
• Server-Sent Events stream for live status updates.
• MongoDB model for device registration and activation state.
• API routes for user status and registered meeting devices.
• Secure headers and content security policy through Helmet.

Tech Stack

• Node.js
• Express
• Zoom Apps SDK
• Python
• Flask
• MongoDB / Mongoose
• Server-Sent Events

Architecture

• app.js - Express app setup, security middleware, Zoom routes, API routes, and proxying to the Python backend.
• server/ - Zoom auth, views, routes, controllers, helpers, and MongoDB models.
• voice_server.py - Flask service that simulates realtime voice-authentication probabilities and exposes JSON/SSE endpoints.
• public/ - Front-end assets served by the Zoom App.

Run Locally

Install Node dependencies:

npm install

Create environment variables from .env.sample, then start the Python status service:

pip install flask flask-cors
python voice_server.py

Start the Zoom App server:

npm run dev

Required Environment

• ZM_CLIENT_ID
• ZM_CLIENT_SECRET
• ZM_REDIRECT_URL
• SESSION_SECRET
• MONGO_URI
• PYTHON_BACKEND_URL (defaults to http://localhost:5000)

Project Notes

This repository is a prototype system for AI-assisted meeting security. The current Python service simulates realtime probabilities, making the application ready for replacement with a trained voice-authentication model.