EbroBot: Your Intelligent LLM Chatbot

Overview

EbroBot is a powerful, interactive chatbot application built using state-of-the-art language models. It incorporates the capabilities of OpenAI's GPT-4o-mini and Chroma vector database to deliver context-aware, high-quality responses while leveraging uploaded documents as a knowledge base. The project is designed to simplify human-computer interaction through natural language processing and knowledge retrieval.

---

Project Structure

1. chatbot.py

This script serves as the main entry point for the chatbot interface. It:

• Configures and initializes the chatbot's language model (ChatOpenAI) and vector database (Chroma).
• Sets up file upload functionality to extend the chatbot's knowledge base dynamically.
• Implements a Gradio-based user interface for seamless interaction.

2. ingest_database.py

This script prepares the knowledge base by:

• Loading documents (e.g., PDFs) from the data directory.
• Splitting the documents into manageable chunks for efficient processing.
• Embedding the chunks using OpenAIEmbeddings and storing them in the Chroma vector database.

---

Features

1. Chatbot with Enhanced Retrieval

• Streamed Responses: Generates partial answers while processing longer queries for a real-time experience.
• Context-Aware Conversations: Leverages conversation history to provide coherent responses.
• Knowledge-Driven Answers: Responds based solely on the knowledge provided by the uploaded documents.

2. File Upload

• Users can upload text-based files to dynamically enhance the chatbot's knowledge base.
• Documents are processed, embedded, and stored in the vector database for retrieval.

3. Document Ingestion

• Handles large datasets with efficient chunking and embedding strategies.
• Ensures persistence of the vector database for continuous use.

---

Prerequisites

1. Environment Setup

• Python 3.8+
• Install dependencies using:

  pip install -r requirements.txt

2. API Keys

• Ensure that the OpenAI API key is set up in a .env file:

  OPENAI_API_KEY=your_openai_api_key
  

---

Usage

1. Ingest Documents

To process and store documents in the vector database:

python ingest_database.py

2. Launch the Chatbot

Start the Gradio-based chatbot interface:

python chatbot.py

The chatbot will be accessible at the URL provided by Gradio (e.g., http://127.0.0.1:****).

---

Directory Structure

.
├── data/                  # Directory for documents to be ingested
├── chroma_db/             # Persistent Chroma vector database
├── chatbot.py             # Main chatbot script
├── ingest_database.py     # Script to ingest documents
├── .env                   # Environment variables
├── requirements.txt       # Python dependencies

---

Dependencies

• LangChain
• OpenAI Python SDK
• Gradio
• Chroma
• Python-dotenv

---

Future Enhancements

• Support for additional file formats (e.g., Word, CSV).
• Advanced conversational memory to handle complex dialogs.
• Deployment on cloud platforms for scalability.

---

Contributing

Contributions are welcome! Please follow the standard GitHub workflow:

1. Fork the repository.
2. Create a feature branch.
3. Commit your changes.
4. Submit a pull request.

---

License

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

---

Acknowledgments

• OpenAI for providing the language model API.
• LangChain for the robust integration framework.
• The open-source community for their contributions.