Inspiration
Inquizitive was born out of a common frustration in education: creating engaging, personalized quizzes is hard and takes way too much time. Whether you're a student trying to study from dense research papers or a teacher building assessments, the process is usually tedious and inefficient. We wanted to make it simple. Imagine dropping in a PDF or a link and instantly getting a high-quality, interactive quiz—powered by AI. That’s the spark that led to Inquizitive.
What it does
Inquizitive turns static content into dynamic learning. With just a few clicks, it creates interactive quizzes using AI. Here’s what it can do:
- PDF-to-Quiz Conversion: Upload any document and get instant, AI-generated multiple-choice questions
- Web-to-Quiz Conversion: Paste a URL and let our smart text extraction do the rest
- Interactive Quiz Experience: Clean, user-friendly interface with real-time scoring and feedback
- Personal Quiz Library: Create an account to save and organize your quizzes
- Browser Extension: Generate quizzes directly from any webpage in one click
How we built it
We built Inquizitive using a modern, type-safe full-stack setup:
- Frontend: Next.js 15 + React 19 with TypeScript and Tailwind CSS
- Backend: tRPC for fully type-safe communication between frontend and backend
- AI Integration: Google Gemini 2.0 Flash for generating structured, reliable quiz content
- Database: PostgreSQL with Prisma ORM to manage user data securely and efficiently
- Authentication: Better Auth to handle secure logins and user sessions
- File Handling: AWS S3 for PDF storage, and Cheerio for clean HTML parsing
- Browser Extension: Built with Chrome extension APIs for smooth webpage integration
Challenges we ran into
- AI Consistency: Making AI responses reliable was tricky—we solved it using strict JSON schemas and solid error handling
- Content Extraction: Parsing HTML and converting PDFs into usable text required a lot of fine-tuning with Cheerio and base64 encoding
- User Experience: We focused on making the interface intuitive, with clear loading states and helpful error messages
- Security: We scoped database queries to each user and protected all sensitive procedures
- File Uploads: To handle big files smoothly, we used AWS S3 with presigned URLs for efficient uploads
Accomplishments that we're proud of
- Full Type Safety: Everything is type-safe—from the database all the way to the frontend
- Reliable AI Integration: Gemini 2.0 works like a charm with structured output and consistent results
- Smooth UX: Users get a powerful tool that feels simple and effortless to use
- One-Click Browser Extension: Instant quiz generation from any webpage—no copy-pasting required
- Scalable Architecture: Built to grow with strong security and clean authentication
- Versatile Content Pipeline: Works well with both PDFs and web content, thanks to our processing pipeline
What we learned
- Working with AI: Structured prompts and response validation are absolutely essential for reliable AI features
- React and Next.js: We dove deep into React 19 and Next.js 15, learning tons of performance and optimization tricks
- Type-Safe APIs: tRPC gave us a whole new appreciation for end-to-end type safety
- Cloud Storage: Learned the ins and outs of file uploads and presigned URLs with AWS S3
- Browser Extensions: Gained solid experience with Chrome APIs and how to handle cross-origin issues
- Database Design: Learned how to properly model data and keep user info securely isolated
What's next for Inquizitive
- Smarter AI: Integrate more models to handle different content types and question formats
- Collaboration Tools: Add features for sharing quizzes, group study sessions, and team workspaces
- Deeper Analytics: Help users track progress and get personalized learning tips
- Mobile App: Build native iOS and Android versions for learning on the go
- Enterprise Features: Add support for SSO, bulk uploads, and LMS integration
- More Content Types: Bring in support for videos, slide decks, audio, and structured data
- Gamification: Make learning fun with badges, leaderboards, and social features
Built With
- amazon-web-services
- gemini
- nextjs
- react
- s3
- tailwind
- trpc
- vercel
Log in or sign up for Devpost to join the conversation.