Data Architecture Diagram

Inspiration

We noticed a major gap in how students share knowledge. While we all have skills we’re proud of and others we want to learn, there’s no structured way to connect learners and teachers within the same community. That's how Peer Pop was born — a smart, AI-powered platform that connects students based on the skills they can teach and want to learn.

What It Does

Peer Pop helps students:

Create skill profiles by listing what they know and want to learn
Match with peers who complement their learning goals
Receive email notifications upon being matched
Rate their learning sessions and track progress
Visualize top skills, trending technologies, and peer learning activity in real-time

How We Built It

We built Peer Pop using a full-stack approach:

Frontend: React.js with Tailwind CSS for a clean, responsive UI, Leaflet.js for skill-to-peer map visualization Backend: Node.js + Express for user and skill management APIs, MongoDB to store user data, skill relationships, and session logs MI & Data: Python-based collaborative filtering model using implicit, Recommendations for “Skills You Might Like to Learn”, Streamlit dashboard for live admin and analytics view

Features

Skill-based Peer Matching Instant Notifications on Match Skill Recommender Engine (Collaborative Filtering) Streamlit Dashboard: View Top Skills, Active Learners, Skill Trends Map-Based Teacher Locator by Skill

Challenges We Ran Into

Designing a clean and scalable matching algorithm that avoids reciprocal matches.
Implementing collaborative filtering with implicit feedback from skill presence.
Syncing frontend React logic with skill recommendation APIs in real-time.
Creating realistic synthetic data to train and test our recommendation engine.

Accomplishments We’re Proud Of

-Built a fully working recommendation system without needing explicit ratings -Designed a sleek, interactive UI for students to explore and connect -Used real AI techniques (matrix factorization) to generate valuable insights -Created a data dashboard that can scale to support thousands of users

What’s Next for Peer Pop

-Add in-session scheduling & chat -Integrate with student calendars -Expand to support club-based or class-specific learning hubs -Offer badges and gamified incentives for top contributors -Launch a mobile version

What We Learned

-Working on Peer Pop taught us so much more than just writing clean code. We learned how to:

Design with empathy — understanding real student challenges around peer learning, confidence gaps, and access to help

Turn data into insight — by using collaborative filtering to generate smart, personalized skill suggestions based on peer behavior

Balance scalability and usability — creating a matching algorithm that’s fast, fair, and meaningful for diverse users

Integrate multiple technologies — from MongoDB and Node.js to React, Python, and Streamlit — in a seamless full-stack workflow

Visualize engagement trends — using dashboards to understand what skills are popular, which users are active, and where demand-supply gaps lie

Simulate real-world data — generating a robust synthetic dataset to test ML models in the absence of real user input

But most importantly, we learned that the future of learning is collaborative — and building tools that empower students to learn from one another can be incredibly rewarding.