Case Study: Robinhood cuts logging costs 45% with WarpStream

Robinhood Swaps Kafka for WarpStream to Tame Logging Workloads and Costs

Robinhood, a financial services company with over 14 million monthly active users, faced significant challenges managing the cost and complexity of its massive logging workloads powered by Apache Kafka. Their platform's highly cyclical traffic, tied to U.S. market hours, resulted in substantial waste from over-provisioned compute and storage. The team sought an elastic and cost-effective solution to replace their Kafka infrastructure and turned to vendor WarpStream for its cloud-native streaming data platform.

By implementing WarpStream's diskless architecture built on object storage, Robinhood eliminated the need for complex networking like VPC peering and AWS PrivateLink through the use of Agent Groups, which also solved noisy neighbor problems. The solution's auto-scaling capability kept clusters right-sized for their variable traffic. As a result, WarpStream simplified operations and reduced Robinhood's overall logging costs by 45%, with compute costs down 36% and storage costs reduced by 13%.

Robinhood

Renan Rueda

Software Engineer