With pre-aggregations becoming increasingly important for developers to make data apps fast, responsive, and cost-effective, Cube provides an open-source API layer and a fully managed platform that runs and scales Cube apps in production. Pre-aggregations are condensed versions of the source data, materialized ahead of time, and persisted as tables separately from the raw data, reducing query execution times by orders of magnitude. To store pre-aggregations, developers can use Cube Store, which is a purpose-built and performant storage layer. The process for creating and managing pre-aggregations varies between self-hosted Cube and Cube Cloud, with Cube Cloud offering managed infrastructure and simplified setup. Optimizing pre-aggregation build times involves parameters such as partitionGranularity and refreshKey, while tracking build history over time can be done using custom log monitoring agents or Cube Store's Build History tab. Previewing pre-aggregation data is also crucial, and Cube Cloud provides a dedicated Preview tab for this purpose. Finally, serving requests solely from pre-aggregations requires a Cube cluster configured with a refresh worker and store cluster, but offers significant latency benefits.