Data Documentation, Modeling, Cataloging, Definitions, Descriptions, Ownership (Business, Technical, etc…)
It would be great to have data documentation, data modeling, cataloging, definitions, descriptions and ownership tied to the physical data tables, reports, notebooks, dataflows, semantic models, objects, etc...
Most people agree that doing this is important and generally a missing piece in the data world. If it could be integrated to the physical objects people will use it more.
We could track changes to the documentation, see which objects are missing it, etc...
This could be used in data exploration helping people understand their data more.
Copilot would be a great fit to help with this.
Even just a simple notes link to each object would be a start.
Could this architecture be created in a way that the same parts could be used for documentation in the rest of the Power Platform too?
Thanks
2 Comments
- Status added:New
ThornKevinAdvocate I
Agree, the semantic model details page should include additional attributes next to title and description as an easy starting point for the PG to make Fabric more useful. This would also allow greater insights into the provided data products and later integration within purview. Below some field ideas for the semantic model detail page: - „Classification“: The nature of the exposed data (source-aligned, aggregate, consumer-aligned) - „Ownership“: Data Product Owner, e.g. a specific department within the enterprise - „Security Tags“: Does this data product contain Personally Identifiable Information (PII)? - „Source Systems“: Maybe this could be handled as tag list or be auto-populated. Some of our users just want to search for Siemens Teamcenter (PLM) data within Fabric
