Overview

QuizForge is an end-to-end AI-powered assessment platform that transforms raw study materials - PDFs, PPTs, and documents - into structured, topic-wise timed MCQ quizzes with instant evaluation and detailed explanations.

The platform features a robust document processing pipeline for text extraction and AI-driven question generation using the Google Gemini API. Users can configure quiz settings including difficulty level, number of questions, and time limits.

Real-time scoring, performance analytics, and revision-focused feedback create a personalized prep experience that adapts to each student's learning patterns.

Highlights

  • Supports multiple document formats: PDF, PPT, and plain text
  • AI-driven question generation with configurable difficulty and topics
  • Real-time scoring with performance analytics and revision suggestions

Challenges

  • Building a reliable document parsing pipeline that handles diverse PDF layouts and PPT structures without losing semantic context.
  • Prompt engineering the Gemini API to generate high-quality, unambiguous MCQs with plausible distractors consistently.
  • Implementing real-time quiz timing and scoring that remains accurate across network latency and browser tab switching.

Key Learnings

  • Deepened understanding of LLM prompt engineering for structured output generation.
  • Learned how to architect full-stack Next.js applications with complex async data flows.
  • Gained experience integrating third-party AI APIs with rate limiting and error handling strategies.