Skip to main content
  1. Public practice questions
  2. AB-620 Practice Exams | AI Agent Builder Associate
  3. AB-620 Practice Exam 2

Question 59 — AB-620 Practice Exam 2

An Azure AI Search index contains compliance documents. Users often phrase questions differently from the terminology used in those documents, and the team wants better ranking of natural-language search results. The Azure AI Search service tier supports semantic ranker. What must be done before the maker can activate semantic ranker for the Copilot Studio knowledge source? (Select one answer)

  1. Enable semantic ranker in Azure AI Search.

  2. Add a second vector index to the same connection.

  3. Replace integrated vectorization with Access Key authentication.

  4. Create a metadata_storage_path field for every document.

Correct answers

  1. Enable semantic ranker in Azure AI Search.

Explanations

Correct:

Enable semantic ranker in Azure AI Search.

Semantic ranker must first be available for the Azure AI Search service tier and enabled in Azure AI Search. The maker can then activate it when configuring the knowledge source in Copilot Studio.

Wrong:

Add a second vector index to the same connection.

Semantic ranker does not require a second index, and a connection supports only one index.

Replace integrated vectorization with Access Key authentication.

Authentication method and semantic ranking are separate concerns. Access Key authentication is not a prerequisite for semantic ranker.

Create a metadata_storage_path field for every document.

metadata_storage_path can supply citation URLs, but it does not enable semantic ranking.