Bringing real-time, on-demand customer service to more than 8 million public transit patrons

Bringing real-time, on-demand customer service to more than 8 million public transit patrons


The customer, a public transport company, wanted to determine the dialog flow of its AI chatbot to help patrons with smart card self-service. This project’s prerequisite was to extend the reach of their customer service offerings while providing them an opportunity to streamline and reduce their customer service representatives’ headcount and ultimately reduce cost. 

Also, they wanted to modify the API library to support transactions initially built for an antiquated IVR system.  

They were challenged with the dense technology and development calendar, which needed incredible coordination with teams and SMEs across the US and European countries. 


Neal Analytics created a transaction bot to support balance inquiry, add stored balance, and pass purchases. Our solution included Azure App Services along with associated services like LUIS using Natural Language Processing (NLP) to interpret intents and evoke a proper response in the context of a transaction. Along with this, we used Application Insights to capture telemetry and Power BI to display data visualizations. 

The Neal Analytics team added functionality to report a lost or stolen card, transfer balance, and issue a replacement through the chatbot. We also provided the ability to assist patrons in self-diagnosing any account access issues.  

Finally, we created a secondary bot to help customers plan their trip when using a variety of transportation methods such as walking, biking, and riding the bus. 


Using Neal’s solution, over 8 million people were able to receive chatbot support. The automated chatbot helped reduce strain on customer service representatives. In addition, the public transport company was able to make transaction estimates based on the rider trends.