What is Whatsapp Mcp Voice Agent

whatsapp-mcp-voice-agent is a voice assistant that integrates with WhatsApp using technologies like LiveKit, Cartesia, OpenAI, Silero, and a Multi-Content Processor (MCP).

Use cases

Use cases include automated customer service responses, interactive voice-based applications, and enhancing user engagement through voice commands on WhatsApp.

How to use

To use whatsapp-mcp-voice-agent, clone the repository, set up a virtual environment, install dependencies, configure environment variables, and set up the WhatsApp MCP server according to the provided instructions.

Key features

Key features include voice interaction capabilities, integration with LiveKit for real-time communication, Text-to-Speech functionality via Cartesia, and Speech-to-Text and LLM operations using OpenAI.

Where to use

whatsapp-mcp-voice-agent can be used in customer support, virtual assistants, and any application requiring voice interaction through WhatsApp.

Voice Assistant

A simple voice assistant using LiveKit, Cartesia, OpenAI, Silero, and a Multi-Content Processor (MCP) for WhatsApp integration.

Prerequisites

• Python >=3.13
• uv (Python packager and virtual environment manager)

Setup

1. Clone the repository:

   git clone <your-repo-url>
   cd voice-assistant
   

2. Create and activate a virtual environment (using uv):

   uv venv
   source .venv/bin/activate  # On Linux/macOS
   # .\.venv\Scripts\activate    # On Windows
   

3. Install dependencies:

   uv sync
   

4. Set up Environment Variables:

   Create a file named .env in the root of the project and add the following environment variables with your actual values.
   You can copy the example below:

   LIVEKIT_URL=
   LIVEKIT_API_KEY=
   LIVEKIT_API_SECRET=
   CARTESIA_API_KEY=
   OPENAI_API_KEY=
   

   • LIVEKIT_URL: Your LiveKit server URL.
   • LIVEKIT_API_KEY: Your LiveKit API key.
   • LIVEKIT_API_SECRET: Your LiveKit API secret.
   • CARTESIA_API_KEY: Your Cartesia API key for Text-to-Speech.
   • OPENAI_API_KEY: Your OpenAI API key for Speech-to-Text and LLM operations.

5. Configure WhatsApp MCP (Multi-Content Processor):

   This project uses an MCP to interact with WhatsApp. You need to have your WhatsApp MCP server running and configured.

   Update the mcp_config.json file in the root of this project with the correct paths for your MCP setup:

   {
     "mcpServers": {
       "whatsapp": {
         "command": "{{PATH_TO_UV}}",
         "args": [
           "--directory",
           "{{PATH_TO_SRC}}/whatsapp-mcp/whatsapp-mcp-server",
           "run",
           "main.py"
         ]
       }
     }
   }
   

   • Replace {{PATH_TO_UV}} with the absolute path to your uv executable. You can find this by running which uv (on Linux/macOS) or where uv (on Windows) in your terminal.
   • Replace {{PATH_TO_SRC}} with the absolute path to the directory containing your whatsapp-mcp project. For example, if your whatsapp-mcp project is in /Users/me/dev/whatsapp-mcp, then this value should be /Users/me/dev. The final path in the config will be /Users/me/dev/whatsapp-mcp/whatsapp-mcp-server.

Running the Assistant

1. (Optional) Download necessary model files (e.g., for Silero VAD):
   The voice.py script might have a command or automatically download necessary models on first run. The example below shows a hypothetical command if your voice.py supported it.

   # If your voice.py has a download command, e.g.:
   # uv run python voice.py download-files
   

   Currently, the voice.py script uses livekit.agents.cli.run_app, which typically handles its own setup for plugins like Silero.

2. Start the voice assistant:

   uv run voice-assistant-cli
   

   This uses the voice-assistant-cli script defined in pyproject.toml.

Connecting to the Playground

Once the assistant is running, you can load the LiveKit Agents Playground in your browser and connect to your local agent.

Note: This README assumes you have uv installed and configured in your system PATH.“# Whatsapp-mcp-voice-agent”