Power BI Dataflows are great: fast and reusable! But who tried to implement some of the best practices described in the documentation might have lost some hair due to non-trivial error messages.

There are some hidden traps that you might fall if you try to implement a chain of Dataflows in the same workspace or replicate the incremental refresh that works on the dataset but not in Dataflows.

In this session we'll move ETL steps created in a Power BI Desktop to an architecture using a chain of Dataflows.

We'll discuss pros and cons of different architectures of the chain of Dataflows depending on its sources or update frequency.

In the end, we'll learn how to successfully set up incremental refresh to a table that doesn't have a datetime column.