Inspiration
Studying can be boring, overwhelming, and ineffective—especially when notes, PDFs, and slides feel like walls of text. We wanted to turn studying into an experience: one where learning feels like a story, questions are fun, and knowledge actually sticks. The idea came from noticing how interactive stories and games engage the brain better than traditional textbooks, and we asked ourselves:
“What if every school document could become a story world you could explore, understand, and interact with?”
What it does
UniStory AI transforms your academic materials into living, interactive stories.
Upload PDFs, DOCX, PPTs, or images (including handwritten notes).
Each subject is isolated—your Biology notes won’t mix with Physics.
The AI (powered by Gemini) turns each section into a story with characters, events, and metaphors, while keeping the content accurate.
After each section, the AI generates interactive quizzes and explains answers through the story, making learning intuitive and memorable.
Listen to your stories with voice narration, making it feel like an audio-visual learning adventure.
How we built it
We built UniStory AI as a full-stack Next.js application with a professional, responsive UI:
Frontend: Next.js + TypeScript + Tailwind CSS for a clean, modern interface.
Backend: Prisma + PostgreSQL for structured data and isolated subjects.
AI Integration: Gemini AI for story generation, question creation, and adaptive explanations.
File Processing: PDF parsing + OCR for images to extract content.
Features: Multi-subject isolation, adaptive learning loops, and interactive quizzes.
Every document uploaded is processed section by section, creating a story world that’s unique for each user and subject.
Challenges we ran into
Document parsing & OCR: Extracting meaningful text from PDFs and images while maintaining context was tricky.
AI accuracy vs creativity: Gemini could generate engaging stories, but we had to carefully prompt it to avoid hallucinations and stay true to the content.
Multi-subject isolation: Ensuring subjects don’t bleed into each other required careful database design and AI context handling.
UX for learning: Designing a UI that feels playful yet professional for both students and judges was a delicate balance.
Accomplishments that we're proud of
A fully working MVP that accepts multiple file formats and turns them into interactive, story-based learning experiences.
AI-generated stories that are fun, memorable, and accurate, with adaptive quiz explanations.
Clean, professional, judge-ready UI that’s responsive and visually engaging.
Successfully implemented subject isolation, ensuring organized learning for every student.
What we learned
How to blend AI with education effectively, balancing engagement with accuracy.
Advanced Next.js + Prisma + Gemini AI integration for real-world document processing.
Designing interactive educational flows that are both fun and effective.
Handling complex UX challenges: multi-format uploads, quizzes, and adaptive learning loops.
What's next for UniStory AI
Add image-to-story diagrams: AI turns diagrams and charts into animated story characters.
Include voice interaction: users can answer quizzes verbally and receive story-based feedback.
Gamification: XP, streaks, and progression to make learning addictive.
Expand AI personalization: AI adapts to learning style and weak areas per subject.
Mobile app version for on-the-go learning.
UniStory AI is just the beginning — our goal is to turn every student’s notes into a personal, interactive learning universe.
Built With
- gemini
- javascript
- postgresql
- react
- tailwind
- typescript
- zustand
Log in or sign up for Devpost to join the conversation.