AI Agent platform for online meeting agents with Google Vertex AI, Datadog and Elevenlabs integration.

Gemini Meet: Autonomous Meeting Agents Platform

Gemini Meet is an open-source platform for deploying autonomous AI agents into video calls (Google Meet, Zoom, Teams).

Unlike simple transcription bots, Gemini Meet agents are active participants: they can speak, use tools (via MCP), and reason in real-time using Google's Gemini models (or other providers).

Gemini Meet

🚀 The Platform

The platform consists of two parts:

1. The Dashboard: A web interface to manage agents, configure tools (MCP), and view live meeting details.
2. The Agents: Ephemeral Docker containers spawned by the dashboard that actually join the meetings.

⚡️ Quickstart

The recommended way to use Gemini Meet is through the Dashboard.

Prerequisites

• Docker (Must be running)
• Node.js 18+ & npm
• PostgreSQL (Or use the docker-compose in dashboard/ directory)
• Google Cloud Credentials (For Gemini)

1. Build Agent Images

The dashboard needs these images to spawn agents. You must build them first. Make sure the .env file is present in the root directory.

cp .env.example .env

Fill in the required variables.

Then build the images:

# 1. Build the base image
docker build -f docker/Dockerfile -t ghcr.io/gemini-meet:latest .

# 2. Build the agent image (used by the dashboard)
cd dashboard
docker build -f Dockerfile.agent -t gemini-meet-with-node:latest .
cd ..

2. Start the Dashboard

Navigate to the dashboard directory and follow the setup there.

cd dashboard
cp .env.example .env
# Edit .env with your Google Cloud Credentials and Database URL
npm install
npm run init-db
npm run dev

Detailed Setup: For full database setup and production deployment instructions, see the Dashboard Documentation.

3. Create an Agent

Open [http