Understanding operational AI
The DAM uses AI to do two practical jobs: describe assets and tag them with keywords. Together these power how end users find assets, through descriptive search and keyword filtering. This page explains what the AI produces, how the pieces connect, and why consistent tagging improves findability.
What operational AI does
Operational AI is AI applied to the routine work of describing and tagging assets, rather than creating new content. When run on an asset, it produces two things:
-
A description: a plain-text summary of what the asset contains.
-
Keywords: tags that describe the content, matched to your taxonomy.
Administrators trigger this work through smart tagging. See Tag and edit asset metadata.
Descriptions and keywords
The description and keywords are both metadata stored on the asset.
The description is generated automatically, so no one has to write a summary by hand. Keywords are matched to your own taxonomy rather than a generic, global set, so the tags reflect your brand and content. Because the keyword set is specific to your organisation, the same image can produce different keywords for different organisations.
How AI helps end users find assets
The value of operational AI appears later, when end users search and filter.
Descriptions feed descriptive search, where users search by meaning across the asset name, description, and keywords, and find related terms rather than exact matches. Keywords feed the filter pane, where users narrow the library by selecting relevant tags. See Find and filter assets.
In other words, the AI work done at tagging time is what makes assets findable at search time.
Why consistent tagging matters
Running operational AI across your assets reduces manual effort and produces consistent metadata, which matters most in large libraries where assets are otherwise hard to locate.
Because keywords map to your taxonomy, the quality of that taxonomy shapes the results. A well-structured taxonomy produces keywords that filter cleanly; a sparse one limits how precisely users can narrow results.
It is worth knowing the limits. Descriptive search interprets meaning broadly, so it can surface results that are related but less literal than expected. Keywords are only as useful as the taxonomy they draw on.