Study Masters

An AI-powered study assistant that transforms PDF study material into personalized quizzes with adjustable difficulty, question count, instant feedback, and scoring.

Comment

Inspiration

Students often have study material scattered across PDFs, lecture notes, and other documents. Creating practice questions manually takes time, and it is difficult to quickly turn a large amount of study material into something useful for revision.

We built Study Masters to make this process faster. The idea was simple: upload your study material, let AI understand it, and immediately get a quiz that can be used for active recall and revision.

What it does

Study Masters is an AI-powered quiz generation platform.

Users can:

  • Upload a PDF containing study material
  • Choose the number of questions
  • Choose the difficulty level
  • Generate a quiz using AI
  • Answer multiple-choice questions
  • Immediately see whether an answer is correct
  • View explanations for answers
  • Track their score and percentage
  • Restart the quiz and practice again

The questions are generated from the uploaded study material rather than from a fixed question bank.

How we built it

The frontend was built with React and Vite.

The backend was built with Node.js and Express. When a user uploads a PDF, the backend receives the file using Multer and extracts its text using PDF parsing.

The extracted study material is then sent to the Gemini API with instructions to generate structured multiple-choice questions in JSON format.

The backend validates the generated response and sends the quiz data back to the React frontend, where the questions, answers, feedback, progress, and final score are displayed.

Technology Stack

  • React
  • Vite
  • JavaScript
  • CSS
  • Node.js
  • Express.js
  • Multer
  • PDF parsing
  • Gemini API
  • Git/GitHub

Challenges we faced

One of the main challenges was connecting the frontend and backend so that a PDF uploaded through the website could be processed and turned into usable quiz data.

Another challenge was handling AI-generated responses reliably. The backend expects structured JSON from Gemini, so we added validation and error handling to prevent invalid responses from breaking the quiz.

We also had to handle PDF extraction, file size limits, loading states, API errors, answer validation, quiz state management, and score calculation.

What we learned

Building Study Masters helped us understand how a complete AI-powered web application works rather than just working on an isolated AI feature.

We learned how a React frontend communicates with a backend, how uploaded files can be processed on a server, how APIs can be integrated into an application, and how AI-generated data can be structured and validated before being displayed to users.

Most importantly, we learned how to turn an idea into a working end-to-end application.

Future improvements

We would like to add:

  • Flashcard generation
  • Study history and progress tracking
  • User accounts
  • More question types
  • Support for scanned PDFs using OCR
  • Personalized quizzes based on previous performance
  • Topic-wise quiz generation
  • More detailed learning analytics

Built With

Share this project:

Updates

Submission history