Introduction
After you identify the event log data from the source systems, your next step is to perform required transformations of the data to make it usable for Microsoft Power Automate Process Mining. Not all potential event log data requires transformations. The goal of transformations is to make the data as uniform as possible for use by process mining. For example, this process might include transformations to correct data types, split columns, and create other changes that make the data easier to consume as part of process analysis.
Depending on how you ingest data into process mining, the transformation can take place when you're loading the data, or you can complete the transformation before the loading process. Regardless of which approach you choose, most transformations in this module would still apply. The only difference would be the tools that you use to complete the transformations.
If you're doing the transformations outside of process mining, you can use any tools that you're familiar with. If you do the transformations during the load process, you must use the Dataflow option to ingest data that uses Power Query to perform the transformations.
As part of the data transforming and loading process, you also make decisions around filtering and creating calculated fields. You can complete these tasks during the transform and load processes, or you can complete them later during the process analysis efforts. This module later explores when to use each option.
Often, data transforming and loading isn't a one-time activity. If changes are needed to the transformations, or if you require more recent data, refreshing the process can bring in new data.
The rest of this module explores these topics in more detail.