One of the key values of the Delta Lake format is the ability to track changes over time to the data. With native support for the Delta Lake table format in Customer Insights, you can now take advantage of Delta Lake incremental data updates to accelerate processing in Customer Insights - Data.

It's typical for customer profile data to change only by a few percentages each day. Instead of reprocessing the entire data set every day, Customer Insights - Data can process just the changed source data for unification in the Delta format, saving time and getting you results faster. This improvement happens automatically when all the tables are in the Delta format. No configuration or change is required.

Incremental processing of segments, measures, activities, and predictions is planned to be added in the coming months.

Virtual Network support uses Azure subnet delegation to manage outbound traffic from Power Platform at runtime. Using Azure subnet delegation avoids the need for protected resources to be available over the internet to integrate with Customer Insights - Data.

Customer Insights - Data uses Virtual Network support provided by Power Platform to securely configure Private Link connections to your Azure storage behind a firewall. Customer Insights - Data uses private links for three scenarios:

• Data input: Data sources are used to bring data from your Azure storage into Customer Insights - Data for processing.
• Data output: At setup, you can configure Customer Insights - Data to write all output to your Azure storage.
• Exports: You can configure exports to send specific tables to your Azure storage.

All new private links use Virtual Network support, and existing private links need to be updated.

More details on the platform support can be found at Virtual Network support overview.

A common challenge when working with customer source data is dealing with old and obsolete source data that is ingested with the data you do want. But preprocessing your source data takes additional time and effort.

Now, Customer Insights - Data allows you to apply row filters directly to any source table you ingest, reducing the data scope to just what you need. Filters are applied once on the Tables page where source tables are now displayed on their own tab separated from the output tables created by Customer Insights. Select the row filter link next to any table to simply set the desired filters.

To optimize processing speed, data is filtered in Spark when it's processed by the unification, segments, measures, activities, and models. The full, unfiltered data set is available when downloading it from from the Tables page.

Key values of this feature include:

• Improved processing time: By removing unwanted rows, the system no longer processes data that isn’t needed, which can reduce processing time.
• Improved output data: Removing unnecessary data from the output makes your data easier to use by consumers who will no longer need to exclude unwanted data.

Preview limitation The condition 'Is any of' {list of values} is supported for string data only.