Player 2 - ML-Powered Matchmaking System: Next-Generation Recommendation Engine

2027 · 2027 Competition

School: School of Computer and Information Sciences
Category: Corporate SponsoredPrimary

Project Overview

One Liner: ML-Powered Matchmaking System: Next-Generation Recommendation Engine

Abstract

Launched to early access on iOS and Android in August 2023, PLAYER 2 is a social
networking and team building app for gamers. Currently, our matchmaking system uses a
survey-based compatibility algorithm. While functional for early-stage growth, we're ready to
evolve into a production-grade machine learning recommender system that learns from user
behavior and delivers personalized matches at scale—similar to systems used by Netflix,
Spotify, Hinge, and Bumble.
This project will design and implement a hybrid ML architecture that combines explicit
preferences (survey data) with implicit behavioral feedback (swipe patterns, match outcomes,
messaging engagement) to provide accurate, personalized recommendations while solving the
cold start problem for new users.

No video available.

Screenshots

0 image(s)

No screenshots uploaded yet.

Team Members

David Wang
Lead
Caleb Liang
Zaafir Hasan
Abdul Chaudhry
Mohamed Fadul
Nicolas Plaia

Advisors

Jeff Salvage
Jeff Salvage

Stakeholders

Colin Johnson
Colin Johnson