Case Study: Riot Games achieves 50% faster data processing and a better player experience with Databricks

Riot Games - Customer Case Study

Riot Games, maker of League of Legends with over 100 million monthly active players and more than 26 PB of data (500 billion data points), faced scaling and collaboration challenges from legacy EMR infrastructure. Disjointed workflows, slow and manual ETL, and difficulty monitoring petabytes of streaming network data across hundreds of thousands of configurations made it hard to pinpoint network issues, deliver personalized content, curb abusive behavior, and build the ML models needed to improve the player experience.

By moving to the Databricks Lakehouse Platform and Delta Lake, Riot centralized data access, automated ETL and job scheduling, and enabled collaborative, real-time model development. ETL performance is now 50% faster, powering recommendation engines that boost conversions, real-time anomaly detection that prevents gameplay lag, and deep-learning-based toxicity detection that reduces abusive behavior — together improving network performance, player satisfaction, retention, and lifetime value.

Riot Games

Colin Borys

Data Scientist