Skip to content

README-VIRTUAL-QUIZ.md

Latest commit

 

History

History
98 lines (66 loc) · 2.37 KB

README-VIRTUAL-QUIZ.md

File metadata and controls

98 lines (66 loc) · 2.37 KB

Virtual Quiz System

A system that allows users to upload notes (PDFs or images), extract text, and ask questions to an AI chatbot about the content.

Features

  • Upload PDF documents or images
  • Extract text using OCR (for images) or PDF parsing
  • Ask questions about the document content
  • AI-powered responses using Gemini AI
  • Fallback mechanisms when AI services are unavailable

Setup

Prerequisites

  • Node.js (v14 or higher)
  • npm or yarn
  • Gemini AI API key (already in your .env file)

Installation

  1. Make sure your .env file contains the Gemini API key:
GEMINI_AI=your_gemini_api_key_here
  1. Install dependencies:
# In the root directory
npm install

# In the client directory
cd client
npm install

Running the System

  1. Start the server:
# In the root directory
node start-server.js
  1. In a separate terminal, start the client:
# In the client directory
cd client
npm run dev
  1. Access the application at http://localhost:5175 (or whatever port Vite is using)

Troubleshooting

Connection Issues

If you see "No response from server" or "Failed to connect to server" errors:

  1. Make sure the server is running on port 5000
  2. Check that the Vite proxy is configured correctly in client/vite.config.js
  3. Restart both the server and client

File Upload Issues

If file uploads fail:

  1. Make sure the uploads directory exists in the root of the project
  2. Check that the server has write permissions to this directory
  3. Verify that the file size is under 10MB

Text Extraction Issues

If text extraction fails:

  1. For PDFs, make sure the PDF is not encrypted or password-protected
  2. For images, ensure the image is clear and the text is readable
  3. Try a different file format if possible

How It Works

  1. Upload: Files are uploaded to the server and stored in the uploads directory
  2. Extract: Text is extracted using pdf-parse (for PDFs) or Tesseract.js (for images)
  3. Chat: User questions are sent to the Gemini AI with the extracted text as context
  4. Response: AI generates responses based on the document content

Technologies Used

  • Frontend: React, Vite, Tailwind CSS
  • Backend: Express.js
  • AI: Gemini AI
  • Text Extraction: pdf-parse, Tesseract.js
  • File Storage: Local filesystem