DEV Community

Streaming Audio: A Confluent podcast about Apache Kafka®

Streaming Real-Time Sporting Analytics for World Table Tennis

Reimagining a data architecture to provide real-time data flow for sporting events can be complicated, especially for organizations with as much data as World Table Tennis (WTT). Vatsan Rama (Director of IT, ITTF Group) shares why real-time data is essential in the sporting world and how his team reengineered their data system in 18 months, moving from a solely on-premises infrastructure to a cloud-native data system that uses Confluent Cloud with Apache Kafka® as its central nervous system. 

World Table Tennis is a division within the International Table Tennis Federation (ITTF). It manages events and commercializes its software application for real-time event scoring worldwide. Previously, ITTF scoring was processed manually with a desktop-based, on-venue results system (OVR) —an on-premises solution to process match data that calculated rankings and records, then sent event information to other systems, such as scoreboards. 

To provide match status in real-time, which makes the sport more engaging for fans and adds a competitive edge for players, Vatsan reengineered their OVR system to allow instant data sync between on-premises competition systems with the Cloud. 

The redesign started by establishing an event-driven architecture with Kafka that consolidates all legacy data sources, including records in Excel along with some handwritten forms (some dating back 90 years, even including records from the 1930 World Championship). 

To reduce operational overhead and maintenance, the team decided to stream data through fully managed Kafka as a service on Azure, for a scalable, distributed infrastructure. Vatsan shares that multiple table tennis events can run in parallel globally, and every time an umpire marks scores in a table, the data moves from the venue into Confluent Cloud, and then the score and rankings are sent to betting organizations and individuals on their mobile apps. 

EPISODE LINKS

Episode source