Knowledge Check

1.

An analytics team needs to ensure that a lakehouse containing customer personal data is protected throughout the entire data lineage, including downstream semantic models and reports. Which sensitivity labeling capability automatically propagates the label to dependent items?

Default labeling

Mandatory labeling

Downstream inheritance

2.

A data team wants to designate a semantic model as the authoritative, single source of truth for product reference data across the organization. Which endorsement level is most appropriate?

Promoted

Certified

Master data

3.

A governance lead reviews the OneLake catalog and notices that many items lack descriptions and sensitivity labels. Which tab in the OneLake catalog provides recommended actions to improve governance posture?

Explore tab

Govern tab

Secure tab

4.

An organization deploys a data agent that answers executive questions about quarterly revenue. The agent has access to both a certified semantic model and an unpromoted experimental dataset. How does endorsement affect the agent's behavior?

The agent ignores endorsement metadata and queries both sources equally.

The agent prioritizes the certified semantic model as a more authoritative source.

The agent only queries items that have the master data designation.

5.

An analytics engineer documents a semantic model with the description 'Monthly retail sales revenue by region and product category, refreshed daily from the certified lakehouse.' How does this affect AI agent responses compared to a generic description?

The description has no effect because AI agents rely only on sensitivity labels for context.

The detailed description provides context that helps the AI agent generate specific, accurate responses about the data's scope and refresh frequency.

The description only affects human users browsing the OneLake catalog, not AI agent behavior.

Feedback