Power BI Dataflows: How does the new tool work

Dataflow is a new Power BI service functionality for ETL self-service processes that allows data analysts and/ or business users to share reusable data entities and to build them using Power Query.

In contrast with Power BI datasets which, at the end of the day, are merely Analysis Services table models, Dataflows are stored as entities in the Common Data Service (CMD) in Azure Data Lake Storage Gen2 (ADLSg2). Both the Datasets (Analysis Services), as well as the Dataflows (CMD), are characterized by an underlying technology that is opaque to users.

Parameter Tables in Power Query


The way to make a dynamic Power Query solution is to make sure that it can be easily customizable and, at the same time, be able to work as more of a pattern rather than just a unique one off solution to a single file, table or whatever it is that your query works against. This is exactly where Parameters and Parameter Tables jump in.

We’ve seen before how you can create Parameters using the Manage Parameters feature, but what if you want to have your own parameters coming from a Table from any other source?

The purpose of having Parameters this way is that it’s dynamic, on every refresh it’ll get the latest values, and it gives another end-user interaction layer, where people can input or change values of parameters completely outside of Power Query / Power BI perhaps on an Excel table, or perhaps on its own storage like a SQL Server database.

In the end, it’s all about providing the user with the friendliest of experiences based on their specific requirements.