Web2025

NBA Insights

Big Data + ML Platform

Big Data project for class where a Hadoop cluster was modernized into a BigQuery-based architecture with a frontend-consumable ML model. The app lets users explore player and team statistics, plus a Victory Predictor trained on 600K+ NBA games to estimate whether a team won from player stats.

Role
Data + Frontend Builder
Stack
BigQueryBigQuery MLHadoopReactData Viz
Overview

NBA Insights is a data product that turns a large basketball dataset into an explorable web experience. The project modernizes a Hadoop-based workflow into BigQuery and exposes team, player, and prediction data through a frontend interface.

Focus

The main focus was connecting the data layer with a usable product surface: preparing BigQuery outputs, working with BigQuery ML, and designing a React interface that made statistics and win predictions easier to understand.

Gallery