DOC-Analyser-Using-LLM

📄 AI Document Analyzer

AI-powered Document Analysis Web Application
Leverage the power of Large Language Models (LLMs) to instantly extract insights and answer questions from PDF and TXT documents.

🚀 Overview

The AI Document Analyzer is a Streamlit-based web app that allows users to upload documents and interact with their content using cutting-edge AI models.
Whether you want to analyze reports, research papers, or any text-based content, this tool provides real-time, context-aware answers.

✨ Features

📂 Universal Document Upload – Supports PDF and TXT formats
🤖 Dual AI Backend –
- Ollama (Offline): Run locally without internet
- Gemini (Free Cloud): Google cloud-based AI
⚡ Real-time AI Analysis – Ask questions and get instant answers
🖥 User-Friendly Interface – Clean, intuitive Streamlit UI
🔒 Safe and Lightweight – Minimal setup, optional cloud use

📄 Supported File Types

File Type	Extraction Method
PDF	PyPDF2 library for structured text extraction
TXT	Direct UTF-8 file reading

⚠️ Note: Scanned or image-based PDFs may not extract text correctly.

🤖 AI Models

1️⃣ Ollama (Offline)

Runs locally on your machine
Uses the tinyllama model
Works without internet once installed
Requires Ollama service running

2️⃣ Gemini (Free Cloud)

Uses Google Gemini Flash model (gemini-flash-latest)
Requires internet connection
Requires GEMINI_API_KEY environment variable
Free-tier friendly and stable for small/medium documents

🛠 Installation

Prerequisites

Python 3.7+
pip package manager

Install Dependencies

pip install streamlit PyPDF2 google-genai

Optional: Ollama Setup

Install Ollama: https://ollama.ai/

Pull the required model:

ollama pull tinyllama

Start Ollama service:

ollama serve

Optional: Gemini Setup

Get a free API key: https://makersuite.google.com/app/apikey

Set environment variable:

# Windows
set GEMINI_API_KEY=your_api_key_here

# Linux/Mac
export GEMINI_API_KEY=your_api_key_here

💡 Usage

Start the app:

streamlit run app.py

Open browser at http://localhost:8501

Upload PDF or TXT document
Select AI backend:
- Ollama (Offline) → Local processing
- Gemini (Free Cloud) → Cloud-based processing
Enter your question about the document
Click Analyze Document to get AI-powered answers

🏗 Project Structure

doc-analysier/
├── app.py              # Main Streamlit app
├── README.md           # Documentation
└── .git/               # Git repository files

🔑 Core Functions

Function	Description
`extract_text_from_file()`	Extract text from PDF/TXT
`ask_ollama()`	Query local Ollama LLM
`ask_gemini()`	Query Google Gemini API
`main()`	Handles Streamlit UI and workflow

⚠️ Limitations

PDF extraction may fail on scanned or image-based documents
Ollama tinyllama model has ~3000 character context limit
Gemini API has free-tier rate limits
Very large documents may need truncation

🛠 Troubleshooting

Issue	Solution
“PyPDF2 not installed”	Run `pip install PyPDF2`
“Ollama not found”	Install Ollama and start the service
“GEMINI_API_KEY not found”	Set environment variable correctly
“Unsupported file type”	Upload PDF or TXT only
“PDF text extraction fails”	Check if PDF is image-based or complex

🏗 Development Stack

Python & Streamlit → UI framework
PyPDF2 → PDF text extraction
google-genai → Gemini cloud AI integration
Subprocess → Ollama CLI interaction
Pathlib & Tempfile → File management

📝 Future Enhancements

Retrieval-Augmented Generation (RAG) with vector embeddings
Multi-document support
Summarization and highlighting
Chat-like interface for document Q&A

📜 License

MIT License – Open source and free to use, modify, and redistribute.