Pomodoro

• 

  Completed Assignment/ Actual Time Needed

• 

  Start Date

• 

  Assignment Information

• 

  AI Prediction Section

• 

  Sign Up Page

• 

  Landing Page

Inspiration: We kept lying to ourselves about how long assignments would take and ended up pulling all-nighters. We built this because we were tired of thinking a paper would take 3 hours when it actually took 10

What it does: Assignment Time Predictor uses AI to give you realistic time estimates for academic work based on your assignment details and past patterns. It breaks down the work by phase, tells you when to actually start, and learns from your real completion times to get more accurate over time.

How we built it: We used React and Vite for the frontend, Google Gemini API for AI-powered time estimates, and Supabase to store assignment history and track user patterns. Everything is deployed on Vercel with a custom dark theme designed to stand out from typical AI projects.

Challenges we ran into: Getting the AI to give realistic estimates instead of generic guesses took a lot of prompt engineering. We also struggled with balancing the design between looking professional and not looking like every other AI hackathon project.

Accomplishments that we're proud of: We built something we'll actually use beyond this hackathon. The personalized learning system that adapts to individual work patterns feels genuinely useful rather than just a gimmick.

What we learned: Prompt engineering is way harder than we thought and getting AI to give practical outputs requires serious iteration. We also learned that sometimes simpler features that work perfectly beat complex features that might break.

What's next for Pomodoro: We want to add Google Calendar integration and Canvas LMS support to automatically pull assignment details. A mobile app version would also be huge since students are always on their phones.

Built With

• css
• html
• javascript
• sql