An AI-powered outbound calling assistant built using Twilio, Google Gemini, FastAPI, and WebSockets. It handles real-time voice streaming, converts speech to text, generates intelligent AI responses, and speaks them back to the callerโall during an active outbound call.
- ๐ Real-time voice interaction over phone calls
- ๐ง AI-generated replies using Gemini
- ๐ Speech-to-Text & ๐ฃ๏ธ Text-to-Speech using Twilio ConversationRelay
- ๐ Environment variable handling using
.env
- ๐ค Automated outbound customer support calls & more
๐ ๏ธ How to Run Start ngrok Open a terminal and run:
bash :-
ngrok http 5050 Update .env file (pls check dummy .env file for refrence)
Copy the HTTPS URL from ngrok (e.g., https://abc123.ngrok.io) and paste it into your .env file:
Terminal :-
-
python main.py
"Enter your phone number in the terminal when prompted. The assistant will place a call and interact in real-time."