Inspiration

Inspiration

As a college student in the US, I often find myself overwhelmed by the amount of reading, lecture notes, and exam preparations. Existing study tools are either too expensive or too fragmented. I wanted to build a solution that acts as a personal, AI-powered tutor that can organize my chaotic notes and help me study efficiently. This inspired me to create StudyBuddy AI.

What it does

StudyBuddy AI is an intelligent study companion designed specifically for university students. Users can upload their lecture notes or PDFs, and the app will automatically:

  • Summarize the key concepts.
  • Generate interactive flashcards for spaced repetition.
  • Create customized practice quizzes to test knowledge before exams.

How we built it

We built the frontend using React and Tailwind CSS for a clean, responsive, and modern user interface (Dark mode included!). The backend is powered by Node.js and Express. For the core AI functionalities (summarization and flashcard generation), we integrated the Kiro API to process the text and extract key information. We also used a PostgreSQL database to store user sessions and generated study materials.

Challenges we ran into

  • Token Limits: Processing large PDF lecture notes meant hitting token limits quickly. We had to implement a text chunking strategy to feed the text into the AI model in smaller parts.
  • Prompt Engineering: Getting the AI to consistently output clean JSON for the flashcards was tricky. We spent a lot of time tweaking the system prompts.
  • Time Management: Balancing this project with our midterm exams this week was definitely a challenge, but we managed to pull through!

Accomplishments that we're proud of

We are incredibly proud that we managed to build a fully functional MVP within the hackathon timeframe. The moment the AI successfully read a 10-page essay and generated accurate, well-formatted flashcards in under 5 seconds, we knew we had something great.

What we learned

This was our first time deeply integrating Kiro's tools. We learned a lot about building AI agent workflows and handling asynchronous API calls efficiently. We also learned the importance of focusing on a specific user pain point rather than trying to build a "one-size-fits-all" app.

What's next for StudyBuddy AI

  • Voice Integration: Allow students to record live lectures and have the AI summarize them in real-time.
  • Community Sharing: A feature where students can share their generated flashcards and study guides with classmates.
  • Mobile App: Port the web app to React Native for iOS and Android.

Built With

Share this project:

Updates

Submission history