LLOYD Assistant

A multimodal agent that uses livekit, deepgram and openai to chat with users in real-time.

Installation

git clone https://github.com/simandebvu/multimodal
cd multimodal
pip install -r requirements.txt

Usage

Describe how to use your project. For example:

python agent.py

Features

• Multimodal agent that uses livekit, deepgram and openai to chat with users in real-time
• Uses the livekit-agents library to create a multimodal agent
• Uses the livekit-plugins-deepgram and livekit-plugins-openai libraries to use deepgram and openai respectively
• Uses the dotenv library to manage environment variables
• Uses the requests library to send requests to the livekit server
• Uses the websockets library to connect to the livekit server
• Uses the asyncio library to handle asynchronous tasks
• Uses the json library to handle json data

Dependencies

This project requires:

• Python 3.8+
• OpenAI API
• Deepgram API
• dotenv

Configuration

1. Create a .env file in the root directory.
2. Add your API keys and organization ID:

OPENAI_API_KEY=your_openai_api_key_here
OPENAI_ORGANIZATION_ID=your_organization_id_here
DEEPGRAM_API_KEY=your_deepgram_api_key_here