Summarize-and-Chat-with-PDF

AI-Powered Document Processing

A web application that uses locally hosted LLMs to summarize PDF documents and enable interactive Q&A about their content. Built with Python, Flask, and Socket.IO for real-time updates.

Features

• PDF Summarization: Generate comprehensive, executive, technical, or bullet-point summaries
• Document Q&A: Ask questions about PDF content and get AI-powered answers
• Local Processing: Works with locally hosted LLMs (like Mistral) for privacy
• Session Management: Save and revisit document processing sessions
• Real-time Progress: Track processing with live updates
• Statistics Dashboard: View document metrics and compression ratios

Technology Stack

• Backend: Python, Flask, Socket.IO
• Frontend: HTML5, CSS3, JavaScript
• AI Processing: Local LLM integration (Ollama compatible)
• Database: SQLite for session storage
• Text Processing: pdfplumber, FAISS for vector search

Getting Started

Prerequisites

• Python 3.12+
• Ollama with Mistral (or other local LLM)
• Node.js (for Socket.IO client)

Installation

git clone https://github.com/Natarajan-R/Summarize-and-Chat-with-PDF.git
cd Summarize-and-Chat-with-PDF
pip install -r requirements.txt

Running

python app.py

Open http://localhost:5000 in your browser

Configuration

Edit app.py to:

Change the model name (mistral by default)

Adjust chunking parameters

License

MIT