In this episode of Pragmatic Works’ Introduction to Data Flows series, Austin Libal continues his walkthrough of Azure Data Factory by focusing on one of the most practical and beginner-friendly transformations available in Mapping Data Flows: the Select transformation. Building on the previous lesson, where the source data was imported and data flow debugging was introduced, Austin demonstrates how users can refine and shape data without needing to write SQL or code manually.
Austin Libal begins by reminding viewers that this lesson builds directly on the first video in the series. In that earlier walkthrough, he covered how to turn on data flow debug mode, connect to a dataset, and begin working with source data inside Azure Data Factory. With that groundwork already in place, this episode shifts attention toward transformation logic, specifically how to choose only the columns needed for a cleaner and more useful final result.
The example dataset used in the video contains movie-ranking data pulled from Letterboxd. Austin uses this sample to make the transformation process easy to follow. The dataset includes the following information:
Although all of these columns exist in the raw source, not all of them are useful for the final output. This is where the Select transformation becomes valuable.
Austin Libal explains that the Select transformation acts as a schema modifier. In simple terms, it allows users to decide which columns to keep, which ones to remove, and even what to rename them to. This is very similar to defining a SELECT statement in SQL, but with a graphical user interface instead of handwritten code.
That visual approach is one of the biggest strengths of Azure Data Factory Data Flows. Rather than writing syntax manually, users can point, click, and configure transformations through the interface while previewing results in real time. For many learners and business users, this lowers the barrier to entry and makes data transformation much more approachable.
After adding the Select transformation from the plus icon in the data flow canvas, Austin names the step Remove Unnecessary Columns. This naming convention helps keep the transformation logic readable and organized, especially as data flows grow more complex over time.
He then reviews the source dataset and identifies two columns that are not needed for the intended result:
These columns may exist in the original source, but they do not add value to the output Austin wants to create. By removing their mappings inside the Select transformation, he reduces the dataset from five columns down to three. This makes the dataset easier to work with and more focused on the information that matters.
After removing the unwanted columns, Austin refreshes the Data Preview pane to validate the new result. The preview confirms that only the remaining desired fields are included.
In addition to removing columns, Austin Libal shows that the Select transformation can also rename existing fields. This is another important step in data preparation because clearer column names make reports, downstream transformations, and analytics easier to understand.
He updates the fields as follows:
These naming improvements provide better context for anyone who uses the transformed dataset later. For example, changing “Year” to “Year Released” removes ambiguity and makes the data easier to interpret in future steps.
Once again, Austin refreshes the Data Preview to confirm that the new column names are reflected properly in the output.
Throughout the lesson, Austin highlights how useful Data Flows can be for learners who want to transform data without heavy coding. Azure Data Factory handles much of the processing behind the scenes, allowing users to focus on logic and desired outcomes rather than syntax. This is especially helpful for teams working on data preparation, reporting pipelines, or ETL processes in Azure.
Key takeaways from Austin Libal’s demonstration include:
Austin Libal closes the video by reinforcing that this transformation is one of the foundational building blocks inside Azure Data Factory Data Flows. While the Select transformation is simple, it plays an important role in preparing data for future steps. Cleaning up fields, narrowing down columns, and standardizing names all contribute to more effective downstream analysis.
For anyone learning Azure Data Factory, this episode serves as a strong introduction to practical data transformation. Austin’s step-by-step walkthrough shows that even basic schema changes can have a big impact on the usability of a dataset. As the series continues, viewers can expect to see additional transformations layered onto this same dataset to build more advanced data flow solutions.
Don't forget to check out the Pragmatic Works' on-demand learning platform for more insightful content and training sessions on Azure and other Microsoft applications. Be sure to subscribe to the Pragmatic Works YouTube channel to stay up-to-date on the latest tips and tricks.