StudyPro

"Claude does the flashcards, Gemini does the watching. You do the learning."

Comment

Inspiration

As university students ourselves, we know the pain of finishing a three-hour lecture with pages of notes and no idea where to start revising. Making flashcards manually is time-consuming, and most study tools feel disconnected from the way students actually learn. We wanted to build something that removes the friction entirely — where the gap between "I have notes" and "I'm actively studying" is measured in seconds, not hours.

What it does

StudyPro is an AI-powered academic platform that transforms your lecture notes and uploaded PDFs into ready-to-study flashcard decks. You can paste your notes, upload a PDF slide deck. Claude reads your content and extracts the most testable concepts, pairing each term with a clear, student-friendly definition. During a study session, students flip through cards with a 3D card interface, type their own answers to get instant AI feedback, and rate each card as Easy, Hard, or Missed — feeding a smart queue that keeps the concepts you're struggling with in rotation. Decks are saved to a personal library so your progress persists between sessions. How we built it

Challenges we ran into

For us, this is our first hackathon and implementing many of the techniques took time and lots of errors. We were also adjusting to using git in the terminal which lead to lots of merge conflicts which took time and sanity to check and correct.

Accomplishments that we're proud of

We're proud that the core feature — paste notes, get flashcards, start studying — works end to end and feels genuinely fast and polished. The 3D card flip animation, AI answer checking with encouraging feedback, and the premade deck generation all came together in a way that feels like a real product rather than a hackathon demo. We're also proud of how clean the UI looks given that it was built entirely from scratch without any component libraries.

What we learned

We learned a huge amount about prompt engineering — specifically how to instruct Claude to return structured data reliably and how to tune the system prompt to produce more precise, exam-appropriate flashcard content. We also deepened our understanding of how FastAPI handles async file uploads, how React conditional rendering replaces traditional routing for single-page apps, and how SQLite can serve as a surprisingly capable backend database for a project like this.

What's next for StudyPro

One of the main features we wanted to implement but were unable to due to time and skill was a deep focus mode. This would use computer vision to monitor your face. If it noticed the user starting to look more dejected, it would prompt a break or motivational messages

Built With

Share this project:

Updates