Vapi Integration with Twilio

Welcome to the Vapi Twilio sample project! This guide will walk you through integrating Vapi with your Twilio account.

Getting Started

To set up this project:

1. Create a Python Virtual Environment

   python -m venv .venv
   source .venv/bin/activate
   

2. Set Up Environment Variables

   • Create a .env file in your repository:

     cp .env.example .env
     

   • Add your VAPI_PRIVATE_TOKEN to the .env file.

3. Install Dependencies

   pip install --upgrade pip &&\
   	pip install -r requirements.txt
   

4. Configure Your Twilio Number

   • Go to your Twilio Console and navigate to your active number.
   • In the Configuration section, change the "A call comes in" setting to Webhook.
   • Set the webhook URL to your_ngrok_url/twilio/inbound_call and save the configuration.

5. Run the FastAPI Server

   fastapi dev ./app/main.py   
   

Exposing Localhost to the Internet

To make your local server accessible over the internet, you can use ngrok. Follow these steps:

1. Install ngrok

   • Download ngrok from the official website.
   • Unzip and install ngrok.

2. Expose Localhost

   ngrok http 8000
   

Obtaining Assistant ID and Phone Number ID

1. Assistant ID: After creating an assistant in the VAPI Dashboard, you will receive an assistant ID. Use this ID in your configurations.
2. Phone Number ID: Purchase a Twilio phone number. The ID of this phone number can be used for your purposes.

Conclusion

This sample project demonstrates how to integrate a custom language model with Vapi. By following these steps, developers can create a more versatile and responsive voice assistant tailored to their users' unique needs.

For more help and detailed documentation, please refer to the official Vapi documentation.