Ticketmaster Leverages Confluent to Reduce Development Friction and Boost Machine Learning

With over 500 million tickets sold each year, Ticketmaster is dedicated to connecting fans around the globe to their favorite teams, artists, and events — helping create memories that will last a lifetime.

After 40 years of continuous innovation and technology advancements, Ticketmaster faced hundreds of software systems and components that interacted with each other in different ways and added tons of friction to software development. In an effort to migrate to a microservices architecture as part of a DevOps transformation, Ticketmaster chose Confluent and Apache Kafka to centralize data from all of its systems and enable even faster innovation.

In addition, Ticketmaster is incorporating machine learning to prioritize ordinary customers over users who fraudulently abuse the system by getting priority access to tickets and then reselling them at a higher price. Having a holistic view of all customer activity has enabled Ticketmaster to build machine learning models that combat this type of abuse and react quickly when the malicious users change their strategy.

"Fans care about what seats are still available, how much do they cost, and what packages they can get with the tickets. And the venue wants to know who’s buying the tickets, are we having problems converting people, do we need to change the offering, or do we need to put another tour date on the calendar because this one is super popular. These are all different ways of looking at the same data. So we put all of this into an inventory stream which gets placed in a different data store for the various uses of that data. Confluent and Kafka have allowed us to get to the position where it’s now fairly low-friction for the data science team to roll out new capabilities with our data."

Chris Smith, VP of Engineering Data Science, Ticketmaster

Learn More About Ticketmaster