In a Foundry pipeline, what are sources?

• A. Processing steps that transform data.
• B. Output datasets produced by the pipeline.
• C. Orchestration mechanisms for scheduling.
• D. Input datasets that feed the pipeline.

Answer: (D)

In a Foundry pipeline, sources are the input datasets that feed the pipeline. They represent the starting data the pipeline reads from before any transformations are applied. You can have one or multiple sources, pulling data from different parts of the catalog to be processed together.

This is distinct from the things that happen later: processing steps (transforms) that modify the data, outputs (the datasets the pipeline writes to), and scheduling or orchestration (how and when the pipeline runs). For example, a pipeline might take a source like sales_raw and a source like customers_master, apply a series of transforms to clean and join them, and then write the results to an output such as sales_summary.