-
stop scrolling
-
flow diagram
🌟 Inspiration
Every day, I saw people—including myself—spending hours scrolling through Reels, Shorts, TikTok, and random posts. It was entertaining, but none of that time resulted in real learning or skill growth.
This made me ask a simple question:
“What if scrolling itself could become a learning experience?”
We live in the most information-rich era, yet we learn almost nothing from 99% of the content we consume. That gap inspired LearnOnScroll: an AI that turns passive scrolling into productive micro-learning.
⚡ What it does
LearnOnScroll transforms everyday online scrolling into personalized learning moments.
Using multimodal AI, it:
- Understands videos, posts, and articles the user views
- Extracts key ideas and concepts
- Generates 10–30 second micro-lessons instantly
- Provides simple explanations, examples, and tiny quizzes
- Tracks the user’s skill growth over time
- Converts unproductive browsing into meaningful learning
It turns the world’s biggest habit—scrolling—into the easiest way to build skills.
🛠️ How we built it
We designed LearnOnScroll around three key technologies:
Multimodal AI
- Vision + text + audio analysis
- Extracts concepts from any content (Reels, Shorts, posts, videos)
- Vision + text + audio analysis
Micro-Learning Generator
- Uses LLMs to create bite-sized explanations, examples, and quizzes
- Compresses complex ideas into simple, understandable lessons
- Uses LLMs to create bite-sized explanations, examples, and quizzes
Personalization Engine
- Learns user interests, goals, learning style, and strengths
- Generates a personalized learning feed and skill graph
- Learns user interests, goals, learning style, and strengths
Skill Dashboard
- Tracks concepts learned
- Shows daily productivity score
- Visualizes long-term growth
- Tracks concepts learned
This architecture makes the app efficient, scalable, and easy to adopt.
🚧 Challenges we ran into
- Defining the right scope: Avoiding over-complicated features while keeping innovation strong.
- Privacy considerations: Ensuring content-analysis happens only with permission.
- Multimodal interpretation: Making AI understand different content formats consistently.
- Lesson consistency: Ensuring micro-lessons are simple, accurate, and personalized.
- Balancing productivity vs. user freedom: Keeping the app helpful, not restrictive.
🏆 Accomplishments that we're proud of
- Creating a unique idea not seen in typical ed-tech apps.
- Turning passive scrolling into a positive, productive habit.
- Designing a solution that works for anyone, anywhere, with no extra effort.
- Strong alignment with AI for social good and global skill accessibility.
- Building a concept that is feasible, scalable, and impactful worldwide.
📚 What we learned
- How multimodal AI can understand real-world content.
- The importance of micro-learning for retention.
- How personalization dramatically increases motivation.
- How big global problems can be solved with simple behavior-based design.
- The value of converting existing habits into learning opportunities.
🚀 What's next for LearnOnScroll – Turning Scrolling Into Learning
- Adding voice-based learning capsules.
- AI mentor mode to guide long-term skill paths.
- Offline micro-learning for low-connectivity regions.
- Browser extension for content-to-learning conversion on desktops.
- Partnerships with educational creators and platforms.
- Turning user behavior insights into a personalized lifelong learning map.
LearnOnScroll will continue evolving into a universal layer of intelligent learning—helping billions grow skills effortlessly, right from their everyday scrolling.
Built With
- llm
Log in or sign up for Devpost to join the conversation.