What type of data does a custom transform in Foundry typically handle?

A custom transform in Foundry is primarily designed to handle semi-structured data. Semi-structured data refers to information that does not conform to a rigid structure but still includes tags or markers to separate elements, which allows for some level of organization. This type of data often exists in formats like JSON, XML, or certain types of CSV.

The ability of Foundry's custom transforms to work effectively with semi-structured data allows engineers to parse, manipulate, and analyze datasets that can vary in structure and depth. This is critical for data engineering tasks that require flexibility, as semi-structured data can contain nested elements and varying schemas that need to be handled dynamically.

While structured data refers to clearly defined data models (such as relational databases) and unstructured data refers to data without any discernible structure (such as text documents, images, etc.), custom transforms excel particularly in environments where the structures may be semi-defined. Complex datasets could be addressed by custom transforms, but the specificity of semi-structured data captures its core function within Foundry.