Skip to content
M↓ View as Markdown ↗

Tags

Tags are labels you attach to Models and Glossary terms. They do two jobs:

  • Finding things: group related models and terms so people can browse and filter them in Studio.
  • Controlling access: decide which models and glossary terms an AI Analyst can use, by tag instead of one by one.

Each Workspace has one shared set of tags, used by both models and glossary terms. A tag called finance on a model is the same tag as finance on a glossary term in that Workspace. The same tag name in two different Workspaces is two separate tags.

Metrics don't have tags.

Availability

  • Who can do this: members with the Edit capability in the active Workspace can add and remove tags and set an AI Analyst's tag-based access.

Add or remove tags

You add tags directly in the Tags field of a model or glossary term. There is no separate step to create a tag: typing a new name creates it.

  1. In Studio, open the model (on its Overview tab) or the glossary term.
  2. Click in the Tags field (placeholder Add tag…) and type a tag name.
  3. Press Enter or type a comma to add the tag. To add several at once, type them separated by commas.
  4. To remove a tag, click the X on its chip.
  5. Click Save changes.

On a model, the help text for this field reads: "Labels used to find this model and to control who can access it."

Tags can't be renamed or deleted

Studio has no way to rename or delete a tag. A tag you remove from every model and term still exists and still appears in AI Analyst tag pickers. Choose tag names carefully.

Find items by tag

  • Glossary: in the glossary list, open the filters and choose a Tag (the default is All tags). To find terms that have no tags yet, set Empty field to Empty tags.
  • Models: in the models list, open the filters and set Empty field to Empty tags to find models that have no tags yet.

Use tags to control AI Analyst access

An AI Analyst's model access and glossary access can follow tags. When you tag a model or glossary term later, it automatically joins (or leaves) every AI Analyst whose access references that tag. You don't need to update each analyst.

Model access by tag

  1. In Studio, open AI Analysts and select an analyst.
  2. Open the Models tab.
  3. In Access Mode, click Edit and choose one of:
    • Models with any of these tags: access follows tags, and a new model with a selected tag is included automatically.
    • All models except these tags: everything is accessible except models carrying a selected tag. With no tags selected, every model is accessible.
  4. Under Tags, open Select tags… and choose one or more tags. Each tag shows how many models carry it.
  5. Review the count and list of matched models (for example "3 models matched"), then click Save changes.

The tag-based modes grant whole models. To narrow an analyst to specific dimensions and measures, use Choose specific models instead. See Models in the AI Analyst configuration.

Glossary access by tag

  1. Open the analyst's Glossary tab.
  2. In Access Mode, click Edit and choose one of:
    • All terms: every glossary term is visible.
    • Only terms with these tags: only terms carrying a selected tag are visible.
    • All terms except these tags: every term is visible except those carrying a selected tag.
  3. Choose the tags under Tags, then click Save changes.

If no tags exist yet, the picker shows "No tags exist in this environment yet." (here, "environment" means your Workspace). Add tags to your models or glossary terms first.

Tips for governing with tags

  • Agree on names first. Because tags can't be renamed or deleted, decide on a short, consistent vocabulary (for example finance, sales, pii) before tagging widely.
  • Exclude by tag, but know the limit. You can tag models that hold sensitive data (for example pii) and set general-purpose analysts to All models except these tags with that tag selected. Exclusion is only as complete as your tagging: an untagged model stays accessible in this mode. For analysts that must see only specific models, prefer Models with any of these tags. Tags control which models an AI Analyst can use. They don't replace access controls in your warehouse.
  • Use tags to scope team analysts. Give a team's analyst Models with any of these tags and the team's tag, so new models for that team become available once tagged.
  • Check untagged items. Use the Empty tags filter regularly so new models and terms don't slip past your tag-based access rules.

Tags and version history

Tags are part of a model's and a glossary term's version, so adding or removing tags creates a new version that you can compare and restore. See Version history and restore.