Inspiration
My friend's friend mentioned wanting better book discovery. I realized I hadn't read in months but could scroll Instagram for hours, so I thought - what if book discovery worked like social media?
What it does
PageOne is vertical scrolling for book discovery where users swipe through actual first chapters instead of videos. AI learns from preferences and behavior to recommend personalized books, with features for saving to lists and voting on recommendations.
How we built it
Built with Expo React Native for cross-platform mobile and Supabase for authentication and database. Google Gemini API generates personalized recommendations while Google Books API provides metadata and covers.
Challenges I ran into
Spent a long time debugging environment variables, Expo is picky about spaces in .env files. Hit infinite recursion in Supabase RLS policies and had to use tunnel mode because Purdue's network blocks development ports.
Accomplishments
The AI generates solid personalized recommendations (friends liked it's recommendations).
What I learned
Learned Expo workflow.
What's next
Core concept works, but there's room to add social features like following friend's reading activities, collaborate on reading lists and forming book clubs.
Log in or sign up for Devpost to join the conversation.