Skip to content

README.md

Latest commit

 

History

History
111 lines (72 loc) · 2.51 KB

README.md

File metadata and controls

111 lines (72 loc) · 2.51 KB
Raw

📄 AI PDF Analyst

Upload PDF → AI Summary → Q&A → Insights (Fully Offline Gen-AI Application)

An intelligent PDF analysis system built using Streamlit, FAISS, and local transformer models. It summarizes long PDFs, answers questions using RAG, and extracts insights — all without any paid APIs, making it perfect for students, AI practitioners, and interview portfolios.

🌍 Live Demo (VS Code Public URL)

🔗 https://frrmb4l1-8501.inc1.devtunnels.ms/

✨ Features

📄 PDF Upload & Text Extraction

Extracts full text from any PDF using pdfplumber.

🧠 AI Summary (Local LLM)

Generates high-quality summaries using BART/T5.

🔍 Ask Questions — RAG Powered

Embedding using SentenceTransformers Vector search using FAISS Local model generates answers

📊 Document Insights

Word count, text length, page count, and readability metrics.

🎨 Beautiful & Simple Streamlit UI

Custom CSS, Responsive UI, clean layout, dark theme look.

🖥️ No API Keys Required

Zero cost, zero rate limits, perfect for beginners & demos.

🛠️ Tech Stack

Frontend

  • Streamlit
  • Custom CSS

AI / ML

  • Transformers (T5/BART local models)
  • SentenceTransformers
  • FAISS (Vector Database)
  • Retrieval-Augmented Generation (RAG)

Utilities

  • pdfplumber
  • NumPy
  • PyTorch

📦 Project Structure

AI-PDF-Analyst/

│── app.py

│── requirements.txt

│── README.md

│── assets/

│ └── styles.css

│── modules/

│ ├── pdf_extractor.py

│ ├── summarizer.py

│ ├── rag_engine.py

│ └── insights.py

│── venv/

🚀 Installation Guide

1️⃣ Clone the Repository git clone https://github.com//AI-PDF-Analyst.git cd AI-PDF-Analyst

2️⃣ Create Virtual Environment python -m venv venv venv\Scripts\activate # Windows

3️⃣ Install Dependencies pip install -r requirements.txt

4️⃣ Run the App streamlit run app.py

🧠 How It Works

  1. PDF Processing

    Extract text Split into chunks of 500–700 characters

  2. Embedding + FAISS Index

    Convert chunks into embeddings using SentenceTransformers Store in FAISS for fast similarity search

  3. AI Summary

    Local transformer (T5/BART) generates summary No API needed

  4. RAG Q&A

    Convert question → embedding Retrieve best chunk Local model generates answer