PassabiAI

Inspiration

As a student, I know how hard it can be to find fast, reliable exam prep help — especially without access to a tutor or expensive resources. I wanted to build something that could give students instant, AI-powered explanations and study support, right from their phone.

What it does

PassabiAI is an AI-powered chat app that helps students prep for exams. Students can ask questions, get instant explanations, and study anytime — even offline, thanks to PWA support.

How we built it

PassabiAI is built with JavaScript, HTML, and CSS on the frontend, deployed on Vercel, with Supabase handling authentication and backend data. I added a service worker and manifest for PWA/offline support, and a theme module for dark/light mode.

Challenges we ran into

Authentication was the biggest challenge — getting Supabase auth working reliably took nearly two weeks of debugging. I also had to restructure the entire project (splitting CSS/JS into more files, adding proper assets/data folders) to keep things organized as the app grew.

Accomplishments that we're proud of

Getting a working, deployed AI exam-prep app built entirely from a phone, with no laptop — as a 14-year-old, self-taught developer.

What we learned

Debugging auth issues taught me a lot about how Supabase sessions and tokens work. I also learned how to structure a growing codebase for maintainability, not just quick fixes.

What's next for PassabiAI

Expanding the question bank, adding more exam types, and improving the AI's explanations based on user feedback.

Built With

• css3
• html5
• javascript
• supabase
• typescript
• vercel