How pranavreddygaddam/quizforge works
Project type and purpose
• QuizForge is a full-stack educational web application that uses AI to convert study materials (PDFs or raw text) into interactive quizzes, summaries, and flashcards.
• The frontend is built with Next.js (React + TypeScript + Tailwind CSS). The backend is a FastAPI service (Python) that handles PDF extraction (PyPDF2), AI service calls, and returns structured responses via REST endpoints.
File structure
• Top-level:
• backend/ – FastAPI service
– app/main.py (entrypoint)
– app/models.py (Pydantic schemas)
– app/ai_service.py (OpenRouter AI integration)
– requirements.txt, .env for API key
• frontend/ – Next.js app
– src/app/… (page routes, layout)
– src/components/… (UI components & pages)
– src/lib/api.ts (API wrapper), googleDrive.ts (Google Drive integration), utils.ts
– tailwind.config.ts, next.config.ts, package.json
• No internal database is present; all data flows through REST and AI calls.
• Build and lint config files: eslint.config.mjs, tsconfig.json, postcss.config.mjs
README for additional insights
• Core features: PDF upload, text paste, quiz/summary/flashcard generation, performance analytics
• Backend dependencies: FastAPI, Uvicorn, PyPDF2, OpenRouter API (Qwen 3-32B)
• Documented endpoints: POST /upload-pdf, /generate-summary, /generate-quiz, /generate-flashcards, /check-answers, GET /health
Detail
• Internal folder/module boundaries (e.g., show ai_service.py inside the service layer)
• Data payload types (JSON schemas from Pydantic models)
• Ports/addresses (frontend port 3000, backend port 8000)
• Critical flows: PDF extraction → AI call → response rendering