VoiceCab VoiceAI

Overview

This project implements a voice-based AI taxi assistant using LiveKit agents. The agent helps customers book rides, check ride status, report lost items, and cancel rides. It leverages Google STT, TTS, and LLM services for voice interactions.

Installation

Ensure you have uv installed. Then, install the required dependencies:

uv venv
source .venv/bin/activate
uv sync

Environment Setup

Copy .env.example to .env.local and update the necessary environment variables:

cp .env.example .env.local

Modify .env.local with your API keys and credentials.

Required Credentials

You need to obtain a credentials.json file from Google Cloud Authentication. You can get it from the following link: Google Cloud Service Accounts

Example .env File

Create a .env.local file and include the following variables:

LIVEKIT_API_KEY=""
LIVEKIT_API_SECRET=""
LIVEKIT_URL="wss://navi-app-hlwfg5aq.livekit.cloud"
GOOGLE_API_KEY=""

Running the App

To start the voice agent, use the following command:

uv run main.py dev