Back to projects
AI · UniHack Hackathon — 1st place
Smart Recycling Assistant
Real-time AI waste classification for better recycling.
Problem
Inefficient waste sorting leads to environmental contamination and lower recycling rates, because people often don’t know which bin an item belongs in.
Approach
With the team I built an AI-powered classification app using Python and TensorFlow to identify waste types in real time from an image, behind a Next.js/Node interface with MongoDB storage — a fast build that emphasized teamwork and practical AI.
Outcome
Achieved 95% accuracy in waste identification and won 1st place at a UniHack hackathon.
Tech stack
Next.jsTypeScriptTailwind CSSPythonNode.jsMongoDB
A Python/TensorFlow classification model served behind a Next.js/Node interface, with MongoDB for storage.