A streaming video analytics team ingests billions of events daily into a Unity Catalog-managed Delta...

According to Databricks Delta Lake optimization documentation, Liquid Clustering is a next-generation file organization capability that automatically manages file co-location without requiring explicit partitioning or manual Z-ORDERing. When combined with Predictive Optimization , Databricks automatically maintains clustering across frequently filtered or queried columns, adapting dynamically as query workloads evolve.

This approach eliminates the need for manual maintenance (such as periodic OPTIMIZE or Z-ORDER commands) while improving query performance on large tables—particularly for high-ingest streaming workloads.

Delta caching (B) only improves performance for cached queries and does not address file layout issues, and (D) handles file size optimization but not clustering. Thus, C is the most efficient, modern, and low-maintenance solution recommended by Databricks.