📄 Retrieval-Augmented Generation (RAG) Chatbot

Welcome to the Retrieval-Augmented Generation (RAG) Chatbot repository. This project enables users to upload documents (TXT and PDF), process their content by chunking large files, and interact with that content through a conversational AI interface. The application uses OpenAI’s GPT-4 for generating responses and FAISS for efficient retrieval.

---

🚀 Features

• File Upload: Supports both .txt and .pdf files.
• Multiple Document Handling: Upload and process multiple documents.
• PDF Parsing: Extracts text from PDFs using PyMuPDF with a fallback to PyPDF2.
• Efficient Chunking: Splits large documents into smaller, manageable pieces.
• Conversational Chat Interface: Chat interface for asking questions about the document content.
• Vector Search: Uses FAISS to index and retrieve document chunks.
• Docker & Devcontainer: Pre-configured Docker and VS Code Devcontainer setup for a consistent development environment.

---

📂 Project Structure

RAG-Chatbot/
├── chatbot.py               # Main Streamlit chatbot application
├── Dockerfile               # Docker container configuration
├── docker-compose.yml       # Docker Compose configuration
├── .devcontainer/           # VS Code Devcontainer configuration files
├── requirements.txt         # Python dependencies
├── .env.example             # Sample environment file (with placeholder values)
└── README.md                # This documentation file

---

🛠️ Prerequisites

Before running the application, ensure you have the following installed:

• Docker (Docker Desktop must be running)
• VS Code with the Remote - Containers Extension
• Git
• Python 3.9+
• An OpenAI API Key

---

🔧 Setup Instructions

1. Clone the Repository

Open a terminal and run:

git clone https://github.com/esadkaoui/INFO5940.git
cd INFO5940

---

2. Configure the Environment

1. Create a .env File
   In the root directory, create a file named .env:

touch .env

2. Add Your Environment Variables
   Edit the .env file to include:

OPENAI_API_KEY=your-api-key-here
OPENAI_BASE_URL=https://api.openai.com
TZ=America/New_York

3. Ensure Secrets Are Not Committed
   Your real .env file should be excluded by listing it in your .gitignore. Use the provided .env.example (with placeholder values) to guide users.

---

3. Open in VS Code with Docker/Devcontainer

1. Open VS Code and navigate to the project folder.
2. Open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P) and select:

Remote-Containers: Reopen in Container

3. VS Code will build and open the project inside the container.

---

4. Run the Application

Option A: Inside the Devcontainer

Once inside the container, open a terminal and run:

streamlit run chatbot.py

Then, open your browser and go to:
http://localhost:8501

Option B: Using Docker Compose

1. Ensure Docker Desktop is running.
2. In your terminal, run:

docker-compose up --build

3. Open your browser at the provided URL (typically http://localhost:8501).

---

📜 Usage Instructions

1. Uploading Documents:

• Click "Upload documents" and select one or more .txt or .pdf files.
• The application will process and index the files.

2. Chat Interface:

• Enter your question in the chat input.
• The chatbot uses the uploaded document content to generate contextually relevant responses.
• The final aggregated answer is displayed as a single paragraph.

---

🛠️ Troubleshooting

• PyMuPDF Not Installed Warning:
  If you see a warning about PyMuPDF (fitz) not being installed, install it by running:

  pip install pymupdf

• OpenAI API Authentication Error (401):
  Ensure your .env file contains the correct API key and that all proxy environment variables are unset:

  unset HTTP_PROXY
  unset HTTPS_PROXY
  unset http_proxy
  unset https_proxy

  Also, verify that openai.proxy = None is set in your code.

• Container Issues:
  If the container fails to start:

  docker-compose down
  docker-compose up --build

  Check the container logs for further details.