modellerUpdated 2026-05-04

Data Tags

What this covers

Live Data Tags drawer opened from Model Builder.

Data tags are user-defined labels that group columns by sensitivity, category, or business domain. Tags serve two purposes: they document which columns carry sensitive data (making models auditable), and they drive column-level access control when combined with persona restrictions (see Column-Level Security). This article explains how to create, assign, and manage data tags.

Before you start

What data tags are not

Tags are for access control and governance, not data classification. They tell Tessallite "this column is PII" so the platform can restrict it — they don't tell the data lake "this column is PII" for compliance scanning. Classification belongs at the source; tagging belongs at the semantic layer.

Creating a tag

  1. Open the Data Tags panel in Model Builder (Toolbelt sidebar).
  2. Click Add Tag.
  3. Enter a tag name (e.g. PII, Financial, Internal Only).
  4. Optionally enter a description explaining what the tag means and who should have access.
  5. Click Save.

Assigning columns to a tag

  1. Open a tag from the Data Tags panel.
  2. Click Add Columns.
  3. Select columns from the table/column picker. Columns are listed as table_name.column_name.
  4. Click Save. The tagged columns now show a tag chip in the Table Details drawer.

You can assign the same column to multiple tags (e.g. customers.email could be both PII and Contact Info).

Grouping strategy

Choose a tagging strategy that matches your governance model:

StrategyTagsBest for
By sensitivityPII, Sensitive, Financial, PublicRegulatory compliance (GDPR, HIPAA)
By audienceExternal, Internal, ExecutiveRole-based access in multi-persona models
By domainCustomer, Product, FinanceLarge models with cross-functional teams

Whichever strategy you choose, apply it consistently. Mixed strategies make persona restrictions harder to reason about.

Viewing tags in the model

Tagged columns display a small tag chip next to the column name in the Table Details drawer. The lineage graph shows a "N tagged cols" count on semantic nodes that have tagged columns.

Best practices

Related