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

Question 42 — AB-620 Case Study Set

Based on the scenario, which implementation should Paideffort use for the current regulatory API integration before the planned multi-system release? (Select one answer)

  1. Use an HTTP request node followed by a Set variable node that reshapes the API response with Power Fx.

  2. Use a Power Automate flow that wraps the same single API request even though no additional processing is required.

  3. Use the Foundry compliance agent to retrieve and restructure every regulatory record before grounding the answer.

  4. Use an Azure Function to copy the regulatory records into uploaded knowledge documents before each question.

Correct answers

  1. Use an HTTP request node followed by a Set variable node that reshapes the API response with Power Fx.

Explanations

Correct Answer:

Use an HTTP request node followed by a Set variable node that reshapes the API response with Power Fx.

Explanation:

Paideffort's current integration consists of one straightforward API request whose response can be transformed directly with a Power Fx ForAll expression. There is no pagination, multi-system aggregation, connector requirement, or complex business logic in the current phase, so the direct HTTP-plus-transformation pattern provides the simplest maintainable design.

Wrong Answers:

Use a Power Automate flow that wraps the same single API request even though no additional processing is required.

A flow becomes more appropriate when Paideffort needs multi-source aggregation, complex filtering, connectors, pagination, or reusable business logic. Adding it now introduces another maintenance layer without solving a current requirement.

Use the Foundry compliance agent to retrieve and restructure every regulatory record before grounding the answer.

The retrieval and field mapping are deterministic. Using an agent would add unnecessary model processing and wouldn't improve the simple API-to-table transformation required by the topic.

Use an Azure Function to copy the regulatory records into uploaded knowledge documents before each question.

Paideffort specifically requires frequently changing regulatory information to remain current and supplied dynamically. Converting every request into document ingestion would add unnecessary latency and operational complexity.