In the process of designing a data processing workflow within Databricks, after ingesting raw IoT device data into bronze tables, what is the appropriate next step for storing cleaned and normalized data?