Oct 16, 2026
No image
Real-Time Transit Prediction Platform
Completed

Real-Time Transit Prediction Platform

$100,000+
more 1 year
United States
2-5
view project
Service categories
Service Lines
Artificial Intelligence
Domain focus
Business Services
Subcategories
Artificial Intelligence
AI Automation and Process Optimisation

Challenge

CodeBranch rebuilt the arrival prediction engine and automated the data pipeline for Roadify, a transit technology company in New York whose platform feeds the countdown displays that tell passengers when the next service is arriving.

CodeBranch rebuilt the arrival prediction engine and automated the data pipeline for Roadify, a transit technology company in New York whose platform feeds the countdown displays that tell passengers when the next service is arriving.

Solution

What we built:

  • A new prediction engine that forecasts arrivals down to the second across buses, trains, subways, trams and cable transit.
  • A parallel run of the legacy model, so agencies could compare old and new predictions before switching.
  • An automated data pipeline that validates each agency feed, rejects stale or future-dated data and sends alerts on edge cases, replacing the manual daily publish step

Results

Results:

  • 95% prediction accuracy, up from a baseline too unreliable to measure
  • 700+ transit agencies feeding live arrival data into the platform
  • Zero manual publish clicks

Stack: PHP, SQL, Google Cloud Platform.

What the client says:


"The CodeBranch team was highly responsive and provided real-time updates."
Scott Kolber, CEO, Roadify (verified review on Clutch, 5.0)