AI-102 exam dumps

AI-102 practice question 370 of 493

Designing and Implementing a Microsoft Azure AI Solution. Professional level, Microsoft. Free question with the correct answer and a full explanation.

AI-102 Question 370

Select 4

You are designing a knowledge mining solution using Azure Cognitive Search. You need to process a set of documents to extract custom metadata that is not natively supported by built-in skills, such as detecting specific product codes. You decide to implement a custom skill and include it in a skillset. Which of the following steps are required to integrate the custom skill into the skillset?

  1. A

    Create an Azure Function that implements the custom logic and expose it through an HTTP trigger.

  2. B

    Define an enrichment step in the skillset configuration and reference the custom skill by its HTTP endpoint.

  3. C

    Deploy the custom skill as a pre-trained machine learning model in Azure Machine Learning.

  4. D

    Test the custom skill independently before integrating it into the skillset.

  5. E

    Enable the 'Allow Custom Skills' setting in the skillset configuration.

Show answer and explanation

Correct answers: A, B, D, E

Explanation

To integrate a custom skill into a skillset in Azure Cognitive Search, you need to create the custom logic (commonly using an Azure Function with an HTTP trigger), configure the skillset to reference the custom skill, and enable the 'Allow Custom Skills' setting. Testing the custom skill independently ensures its functionality before integrating. Deploying an ML model is not required unless the custom skill is explicitly based on a machine learning model.

  • A. Correct.

    Correct. Creating an Azure Function with an HTTP trigger is a common way to implement a custom skill since Azure Cognitive Search can call HTTP endpoints.

  • B. Correct.

    Correct. You must define the custom skill as an enrichment step in the skillset and provide the endpoint URL to enable its integration.

  • C. Incorrect.

    Incorrect. While you can use Azure Machine Learning for certain scenarios, deploying a custom skill as an ML model is not necessary for this specific case. The custom skill can be implemented using an Azure Function or another HTTP endpoint.

  • D. Correct.

    Correct. Testing the custom skill independently ensures that it works as expected before integrating it into the skillset, reducing potential errors during deployment.

  • E. Correct.

    Correct. The 'Allow Custom Skills' setting must be enabled in the skillset configuration to permit the use of custom skills during the enrichment process.

Timed practice exam

Take a AI-102 practice test under exam conditions

40 questions in 120 minutes, drawn from this bank, with a score report and a per-question review when you finish.

Start timed exam