Describe the pipeline you want and read the PySpark it would take to build it. It is a fast way to see how sources, transformations, quality checks, and writes fit together.