Music Social Media App/Website (Project Musica)

Project Musica, a fullstack website and app that helps people compile playlists and find music to share with others

Fullstack

Machine Learning

AI

UI

Project Overview

Project Type: Project Musica, a fullstack website and app that helps people compile playlists and find music to share with others
Technologies Used: React Native, PostgreSQL, Python, Javascript, Vercel, Supabase, Huggingface
Timeline: 12 weeks (Spring 2026)
My Role: Lead Designer and Coder

Built a full-stack application that uses a frontend hosted on Expo and a backend hosted on Vercel supports use on Web. This social media app helps user look for songs, create playlists, and connect with other users with similar music tastes. The song database was made with the curation of Spotify and Youtube song across multiple genres, languages, and decades. The search prioritizes our database songs but also populates with Spotify and YouTube results. There is also increased social functionality by computing similarity scores so you can find people who share similar music taste. The homepage recommendation and smart playlist song add relies on our own proprietary ensemble model that uses ALS, Two-Tower, and Spectrogram to have a mixture of experts method of deciding and ranking the best recommendations for users based on whatever is available in the user's playlist or their library + favorited songs.

How to Access:

Live Demo: https://projectmusica.expo.app/

Github:https://github.com/kkim-4/project-musica (currently private)


Ready to build something amazing?

I'd love to connect with you!

Create a free website with Framer, the website builder loved by startups, designers and agencies.