Chat with PDF files using LangChain and Google Gemini-Pro

---

An AI-powered system for extracting, summarizing, and providing real-time Q&A from PDF documents, capable of understanding and responding to user queries.

---

Objectives

The objective of this project is to enable interactive Q&A with multiple PDF files using the Google Gemini Generative AI model. Users can upload PDFs, ask questions related to their content, and receive accurate responses based on the document information.

Demo

Demo-Chat_with_PDF_using_LangChain.mp4

Features

• AI-Powered Extraction: Automatically extracts text from multiple PDFs.
• Summarization: Uses large language models (LLMs) to generate concise document summaries.
• Question Answering: Enables users to query PDF content and receive accurate, context-aware responses.
• Real-Time Interaction: Offers a seamless chat interface for engaging with document content.

System Design

Tools and Libraries Used

• Streamlit: Used for building the web application interface, making it easy to create interactive elements and display information to the user.
• PyPDF2: Used for reading text from PDF files, allowing the app to extract information from uploaded PDFs.
• Langchain: Used for natural language processing tasks, including text splitting, vector embeddings, and question answering.
• Google Generative AI: Specifically, the Gemini model, used for generating answers to questions based on the content of the PDF files.
• FAISS: Used for similarity search, creating a vector store of text chunks from the PDF files to enable efficient retrieval of relevant information.
• dotenv: Used for loading environment variables, allowing the app to securely access the Google Generative AI API key.

Installation

To run the application, first install the required dependencies. Clone the repository first and then intall the requirements.txt file

pip install -r requirements.txt

Usage

Set up your environment variables: Ensure you have access to Google Gemini Pro API and update your .env file with the appropriate API keys. Run the application:

 streamlit run app.py

Upload your PDF documents through the interface, and start querying them using the provided chat interface.