Whatsapp LLM

This project introduce a bot that can help Whatsapp communities (using groups) to enrich their experience by using LLM models.

How does it work?

The project is composed by two main components:

1. A chat-manager that is responsible for connecting to Whatsapp, and interacting with it. This component is written in Go and exposes a gRPC API over a Unix Domain Socket.
2. The bot which is a python application that connects to the chat-manager and uses it to interact with Whatsapp. This component is written in Python and uses the chat-manager to interact with Whatsapp. It is using LangChain and OpenAI.

Running the project

Prerequisites

1. Docker

Configuring the project variable

1. Copy the .env.edit file to .env and edit it to your needs.
   1. You must define the OPENAI_API_KEY variable with your own key
   2. If you wish to use the pre-configured postgres database, there's no need to define it's variable.

Run the project

docker compose up -d

Setting up your bot

1. The first time you connect, the chat-manager container will show you a QR code. Scan it with your Whatsapp mobile application.
2. Add your bot to the group you wish to use it in (using the mobile app, or another Whatsapp web).
3. After the bot is running(both chat-manager and bot containers with no errors), and after the bot got a few messages in some groups, you'll be able to see those groups in the database under the groups table.
4. To allow the bot to interact with the group, you need to set the managed column to true for the group you wish to use the bot in.