RAG-Email-Chatbot

A Retrieval-Augmented Generation (RAG) based chatbot for email interactions. This project implements an intelligent email assistant that can understand and process email content using advanced language models and RAG techniques.

Features

• Email processing and analysis
• RAG-based response generation
• Gmail API integration
• Pinecone based Vector Database
• OPENAI for embeddings and as generator

Example

Prerequisites

• Python 3.8 or higher
• uv package manager (for dependency management)
• Gmail API credentials

Installation

1. First, install uv if you haven't already:

pip install uv

2. Clone the repository:

git clone https://github.com/yourusername/RAG-Email-Chatbot.git
cd RAG-Email-Chatbot

3. Create and activate a virtual environment using uv:

uv venv
.venv/Scripts/activate  # On Windows
source .venv/bin/activate  # On Unix/MacOS

4. Install dependencies using uv:

uv pip install -r requirements.txt

Configuration

1. Create a .env file in the root directory with the following variables:

OPENAI_API_KEY=<your_openai_api_key>
PINECONE_APT_KEY=<your_pinecone_key>
COHERE_API_KEY=<your_cohere_key>

2. Set up Gmail API credentials:

• Go to the Google Cloud Console
• Create a new project or select an existing one
• Enable the Gmail API
• Create credentials (OAuth 2.0 Client ID)
• Download the credentials and save as credentials.json in the project root

Running the Project

1. Ensure your virtual environment is activated:

.venv/Scripts/activate  # On Windows
source .venv/bin/activate  # On Unix/MacOS

2. Run the main application with ingest to upsert emails in vector db:

python ./src/main.py --ingest

3. Run the main application (chatbot mode):

python ./src/main.py

Project Structure

RAG-Email-Chatbot/
├── src/               # Source code
│   ├── __init__.py   # Package initializer
│   ├── main.py       # Main application entry point
│   ├── mail.py       # Email handling and Gmail API integration
│   ├── generator.py  # RAG-based response generation
│   ├── vector_db.py  # Vector database operations
│   └── utils.py      # Utility functions and helpers
├── tests/            # Test files
├── logs/             # Application logs
├── requirements.txt  # Project dependencies
└── .env             # Environment variables

TODO

• Email text cleaning function
• Open source encoder
• Option to change LLM
• Add open source LLMs.
• Code cleanup
• Add testcases using tox.
• Create Docker Image.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributors

_{Sahil Kumar}