Case Study: Prefect achieves real-time workflow observability and lower costs with ClickHouse

Prefect cuts analytics costs to under $8,000/month with ClickHouse

Prefect, an event-driven workflow orchestration platform, faced a challenge scaling its observability features for customers. Their original data stack, using Google BigQuery and PostgreSQL, was not built for real-time analytical workloads. As the volume of workflow events grew to hundreds of millions per day, they encountered performance bottlenecks and rising costs, which prevented them from providing the advanced, interactive analytics their users needed.

By migrating their analytics to ClickHouse Cloud, Prefect gained a database designed for massive volumes of time-oriented event data. The solution enabled new features like Metrics and Automations, providing users with powerful, real-time insights and the ability to trigger actions from workflow events. ClickHouse simplified Prefect's architecture, improved resiliency, and, while not the primary goal, significantly reduced their monthly data stack costs from over $12,000 to under $8,000.

Prefect