Meeting Proxy

README

🎥 Demo

href="https://youtu.be/uFs-ezIrIic"

Click the preview above to watch the full demo.

What is MeetingProxy?

MeetingProxy is an autonomous AI agent that joins Google Meet sessions on your behalf. It listens to the conversation in real time, understands context using an LLM, and responds naturally through a lip-synced digital avatar with a cloned voice — effectively acting as a digital twin in meetings you can't attend. it is basically a fully functional server or plateform which enables the user to achive or make a digital clone with no effort. and it self joins the meeting and itself adjust the mic and camra settings Built as a Final Year Project, MeetingProxy explores the intersection of conversational AI, real-time audio/video pipelines, and browser automation.

Features

Architecture

┌─────────────────┐      ┌──────────────────┐      ┌───────────────────┐
│   React (Vite)   │◄────►│   Flask Backend   │◄────►│   Google Meet      │
│   Frontend UI    │      │   (Orchestrator)  │      │   (via Selenium)   │
└─────────────────┘      └─────────┬─────────┘      └───────────────────┘
                                    │
             ┌──────────────────────┼──────────────────────┐
             ▼                      ▼                       ▼
     ┌───────────────┐     ┌────────────────┐      ┌────────────────┐
     │ Whisper (STT)  │     │ GPT-4o-mini    │      │ ElevenLabs (TTS)│
     │ Transcription  │     │ Reasoning Layer│      │ Voice Synthesis │
     └───────────────┘     └────────────────┘      └───────┬────────┘
                                                             ▼
                                                     ┌────────────────┐
                                                     │  D-ID Avatar   │
                                                     │  Lip Sync Video│
                                                     └───────┬────────┘
                                                             ▼
                                              ┌──────────────────────────┐
                                              │ OBS Studio + VB-Cable    │
                                              │ → Streamed into Meet call │
                                              └──────────────────────────┘

Full diagrams available in /docs/architecture (draw.io source included).

Tech Stack

Getting Started

Prerequisites

• Python 3.10+
• Node.js 18+
• OBS Studio installed with WebSocket plugin enabled
• VB-Audio Virtual Cable installed
• API keys: OpenAI, ElevenLabs, D-ID, Firebase config

Installation

# 1. Clone the repository
git clone https://github.com/mfaaizi/MeetingProxy.git
cd MeetingProxy

# 2. Backend setup
cd backend
python -m venv venv
source venv/bin/activate   # Windows: venv\Scripts\activate
pip install -r requirements.txt

# 3. Frontend setup
cd ../frontend
npm install

# 4. Environment variables
cp .env.example .env
# Fill in: OPENAI_API_KEY, ELEVENLABS_API_KEY, DID_API_KEY, FIREBASE_CONFIG

# 5. Run the app
# Terminal 1 (backend)
cd backend && python app.py

# Terminal 2 (frontend)
cd frontend && npm run dev

The app will be available at http://localhost:5173.

MCP Server Integration

MeetingProxy exposes its core capabilities (join meeting, get transcript, trigger avatar response) as callable tools via a FastMCP server — allowing any MCP-compatible client (like Claude) to control a meeting session programmatically.

cd mcp-server
pip install -r requirements.txt
python server.py

See /mcp-server/README.md for the full tool schema.

Project Structure

MeetingProxy/
├── frontend/          # React + Vite UI (18 page components)
├── backend/           # Flask API + orchestration logic
├── mcp-server/         # FastMCP server exposing MeetingProxy as tools
├── docs/              # FYP documentation, diagrams, test cases
└── README.md

Roadmap

• Multi-platform support (Zoom, MS Teams)
• Multi-language transcription & response
• Persistent memory across recurring meetings
• Meeting summary auto-generation & email digest

Author

Muhammad Faaizi BS Artificial Intelligence · 2022–2026

📌 Focused on Agentic AI, MCP Servers, and Applied ML Systems

GitHub · LinkedIn