PDF-Based Question Answering Chatbot

This project is a chatbot application that answers user questions based on the content of a provided PDF document. If the answer cannot be retrieved from the document, the chatbot provides a fallback response.

Features

• PDF Understanding:
  The chatbot extracts and processes text from the uploaded PDF document and retrieves answers to user queries using the Gemini API.

• Fallback Response:
  When the answer is not found in the PDF, the chatbot responds with:
  "Sorry, I didn’t understand your question. Do you want to connect with a live agent?"

• User Interaction:
  An intuitive web-based interface allows users to upload a PDF, ask questions, and receive responses in real time.

Technologies Used

• Frontend: React, TypeScript, Vite
• Backend: Gemini API (NLP model for question-answering)
• PDF Processing: pdf.js
• Deployment: PDF Chatbot URL

Getting Started

Prerequisites

Ensure you have the following installed:

• Node.js (v14 or higher)
• npm (v6 or higher)

Installation

1. Clone the repository:

   git clone https://github.com/rahulsoni10/PDF-AI-Chatbot.git
   

2. Replace the VITE_GEMINI_API_KEY=your-gemini-api-key in the .env file