Module assessment

1.

An analytics team notices that Copilot in Power BI frequently returns incorrect answers when users ask about revenue. The semantic model has a measure called Rev with no description. What should the team do first to improve Copilot's accuracy for revenue-related questions?

Create a verified answer for every possible revenue question.

Rename the measure to a descriptive name like 'Revenue (USD)' and add a description that explains the business logic.

Mark the semantic model as Approved for Copilot to enable enhanced AI features.

2.

A data team generates an ontology from a published semantic model in Fabric. After generation, several entity types have names like factsales and dimstore. What is the recommended next step?

Delete the generated ontology and rebuild it manually from OneLake tables.

Rename the entity types to business-friendly names like 'SaleEvent' and 'Store' and verify data bindings.

Publish the ontology as-is since the entity types automatically update when the semantic model changes.

3.

A developer configures the AI data schema in the Prep for AI dialog. They hide several surrogate key columns and ETL metadata fields. Which Copilot behavior does this action directly affect?

Copilot no longer includes those fields in its grounding data during preprocessing.

Copilot generates faster responses because it processes fewer tokens.

Copilot automatically creates descriptions for the remaining visible fields.

4.

An organization wants Copilot to consistently return the same visual when users ask 'What were total sales last quarter?' Which Prep for AI feature should the team configure?

AI instructions that define the preferred measure and time filter for sales questions.

Verified answers with trigger phrases that match common ways users ask this question.

Linguistic modeling with synonyms for 'total sales' and 'last quarter.'

5.

An analytics professional builds a well-designed semantic model with clear names, descriptions, and star schema relationships. They want their work to support AI agents beyond Power BI Copilot. Which Fabric IQ capability connects their semantic model to broader enterprise AI?

The Generate Ontology feature that creates entity types, properties, and relationships from the semantic model.

The Approved for Copilot designation that certifies the model for AI consumption.

The Q&A setup that configures synonyms and linguistic relationships for the model.

Feedback