PlayByPlay

The Project

PlayByPlay is an mobile application for sports recommendation and performance prediction using OpenAI and Machine Learning.The app collects input from simple physical tests and user questionnaires, then analyzes the data using OpenAI’s language models to match each user with sports that fit their strengths and circumstances. It also predicts a player’s performance in golf based on their current skill set using machine learning algorithms. This project was inspired by my own experience as a student-athlete. I’ve seen many talented kids never achieve their full potential in sports just because they never discover the ones they’re best suited for. In addition, many athletes struggle to improve and perform big moments. I want to change that and solve both of these issues through this app. One of the most difficult aspects was designing the AI prompts so that they gave accurate, fair, and actionable sport recommendations across many types of users. But by testing iteratively and comparing it with actual dataset from successful athletes, I was able to refine it to the best version.

View this project on YouTube