IceLens

IceLens Screenshot


🏒 IceLens: Full-Stack Hockey Rink Visualization with AI Insights

IceLens is a cutting-edge full-stack hockey analytics platform designed to bring game intelligence to the rink. Built by Corey Yang Smith, IceLens is powered by the Big Data Cup 2021 dataset and offers a live-deployed, open-source solution for coaches, analysts, and sports data enthusiasts.


🚨 Key Features

  • 🏟️ 1:1 NHL Rink Replica: Interactive, rule-accurate rink with event overlays and pass arrows (success/failure).
  • 🎞️ Game Timeline Playback: Scrub, play, and skip through in-game events with live data updates.
  • 📊 Rink & Table Visualizations: Toggle between dynamic rink events and tabular data with filtering by team, player, and event type.
  • 🔥 Heatmaps for Coaches: Visualize shot density, goals, penalties, passes, and more.
  • 🧠 AI Chatbot – PuckQuery: Ask questions like “Who had the most shots?” and get LLM-powered insights via GPT-4.1 Nano + Pandas.
  • 📤 Data Export: Download filtered stats in JSON format for external analysis.
  • ⚙️ Personalized Experience: Persist filters, player settings, and event preferences across views.

🧰 Built With

  • Frontend: React, TailwindCSS
  • Backend: FastAPI, Pandas
  • AI: GPT-4.1 Nano for real-time insights
  • Deployment: Railway

🏁 Ideal For

  • Hockey coaches scouting player performance
  • Analysts conducting game reviews
  • Fans exploring advanced game metrics
  • Researchers looking into sports data visualization