LLM-PDF-AI-Chat

Welcome to LLM-PDF-AI-Chat! 🚀 This application allows users to interact with the content of PDF files conversationally using cutting-edge language models and vector search.

Overview

This project combines Google Gemini's Generative AI, LangChain, and FAISS to extract and process PDF content, enabling users to ask detailed questions and receive precise answers based on the context of the documents.

Features

• PDF Text Extraction: Process multiple PDFs and extract text for analysis.
• Conversational AI: Use advanced AI models to answer questions about your PDFs.
• Embeddings and Vector Search: Efficient text similarity search using Google Generative AI embeddings and FAISS.
• User-Friendly Interface: A clean and intuitive Streamlit app for easy interaction.

Getting Started

Follow these instructions to set up and run the project locally.

Prerequisites

• Python 3.9 or later
• A valid Google Generative AI API key

Setup

1. Clone the Repository:

   git clone https://github.com/your-username/LLM-PDF-AI-Chat.git
   cd LLM-PDF-AI-Chat
   

2. Set Up a Virtual Environment :

bash
Copier le code
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies:

Copier le code
pip install -r requirements.txt

4. Add Your API Key:

Create a .env file in the project directory and add your Google API key:

Copier le code
GOOGLE_API_KEY=your_google_api_key
Run the App:

Copier le code
streamlit run app.py

Usage

Upload your PDF files in the sidebar.
Click on Submit & Process to extract and index the content.
Ask questions in the input box, and the AI will respond with context-based answers.
Example Queries
"Summarize the key findings of the report."
"What are the challenges mentioned in the document?"
"Is there any mention of climate policies in the text?"

File Structure

plaintext
Copier le code
.
├── app1.py                # Main application file
├── requirements.txt      # List of dependencies
├── .env                  # API keys and environment variables
└── README.md             # Documentation

Author

👩‍💻 Chaima Askri 📅 Version 1.0

Feel free to contribute, report issues, or suggest improvements. Happy chatting with your PDFs! ✨