During Loyola’s Spring Hackathon, I built CTA Transit Tracker, a horribly buggy full-stack web app that provides real-time bus and train predictions across Chicago. It integrates CTA’s APIs, GTFS data, and a custom notification system that texts users when their favorite lines are arriving near their home.

This was a major engineering challenge that included:

  • 🔧 Parsing and visualizing GTFS transit data with folium and Leaflet.js
  • 🧠 Creating a notification system with Celery workers that checks for real-time arrivals
  • 📨 Sending alerts via SMS-over-email gateways
  • 🔐 Implementing OTP-based phone authentication
  • 🌐 Building a dynamic dashboard with Flask and Bootstrap

🧠 Technical Highlights:

  • Flask backend handles user auth, prediction APIs, and dynamic routing.
  • Celery with a separate beat scheduler to push notifications based on custom thresholds.
  • GTFS data parsed live from CSVs to render routes and stops.
  • Uses SQLite for fast local data storage, with Flask-Migrate + Alembic for schema changes.
  • Leaflet.js provides an interactive map of buses and trains using emoji and icon markers.
  • Users can set their home location, favorite lines, and receive SMS alerts based on proximity and time-to-arrival.

🧪 Notable Features:

  • 📍 Draggable home marker to set geolocation preferences
  • 🔁 Real-time updates from both CTA bus and train trackers
  • 🗺️ Route and stop visualization using stops.txt and shapes.txt
  • 💬 OTP verification for login without storing passwords
  • 🔔 Background job sends SMS when a favorite line is about to arrive near home

🧑‍💻 Team:

  • Solo project built in under 48 hours
  • Deployed using Gunicorn, Celery, and a Procfile for Heroku-style hosting

🔗 Useful Links: