Simple RAG System for PDF Documents

Welcome to the Simple RAG System repository! This project provides two tools to build and interact with a Retrieval-Augmented Generation (RAG) system for PDF documents using LlamaIndex. Explore and query your PDFs with ease using either a Jupyter Notebook for development or a user-friendly Streamlit app for real-time interaction.

---

📂 Features

• ✅ Upload and Parse PDFs: Process any PDF document for indexing and querying.
• ✅ Ollama Embeddings: Leverage powerful embeddings for accurate content retrieval.
• ✅ Natural Language Queries: Ask questions in plain language and get AI-generated answers.
• ✅ Live Indexing Status: Streamlit app displays real-time indexing progress.
• ✅ LlamaIndex-Powered: Utilizes LlamaIndex for efficient vector storage and retrieval.

---

🔧 Requirements

To get started, install the required dependencies using pip:

pip install streamlit llama-index llama-index-llms-ollama llama-index-embeddings-ollama

---

🚀 Usage

This repository includes two tools for interacting with the RAG system:

🔬 Notebook Version (simple-rag.ipynb)

Ideal for development, experimentation, and step-by-step exploration.

1. Launch the notebook:

   jupyter notebook simple-rag.ipynb

2. Update the PDF file path in the notebook to point to your desired document.

3. Run the cells to index the PDF and interact with the RAG system via queries.

🌐 Streamlit App (simple-rag-streamlit.py)

A user-friendly interface for real-time PDF processing and querying.

1. Run the Streamlit app:

   streamlit run simple-rag-streamlit.py

2. Upload a PDF file through the web interface.

3. Monitor the indexing progress in real-time.

4. Enter a query in natural language and receive an AI-generated response.

---

🧠 Credits

This project is built using the following technologies:

• LlamaIndex: For vector storage and retrieval.
• Streamlit: For the interactive web interface.
• Ollama: For LLM and Embedding models.