Which two features can you use to reduce query execution time for a DirectQuery semantic model?

In the context of a DirectQuery semantic model, user-defined aggregations are crucial for optimizing query execution time. When implemented, user-defined aggregations allow you to create summarized data structures that can be utilized to answer queries more efficiently. Instead of the system running large, detailed queries against the underlying data source, which can be time-consuming, it can leverage the pre-aggregated data for faster response times. This technique reduces the amount of data processed, significantly speeding up query execution.

While automatic aggregation is also beneficial by automatically creating aggregations based on usage patterns, it is the flexibility and control provided by user-defined aggregations that enables tailored optimization for specific query scenarios. Query caching can help with performance but isn't always applicable in DirectQuery scenarios where real-time data is needed, and Direct Lake integration enhances data accessibility but doesn’t directly influence execution times like user-defined aggregations do. Therefore, focusing on user-defined aggregations is a direct method for improving query response times in DirectQuery setups.