When you provision an Azure Document Intelligence resource in the Azure portal, which two items are generated that you use to authenticate API calls?

Connection string and access token

Subscription ID and resource group name

A manufacturing company has a custom Document Intelligence model that extracts part numbers and quantities from work orders. The model performs well on typed work orders but poorly on handwritten ones. They have a small set of handwritten samples. What should they do to improve performance on handwritten documents?

Use the prebuilt read model to pre-process the handwritten text.

Deploy the model and rely on human review for handwritten documents.

Add labeled handwritten samples to the training set and retrain the model.

Increase the resolution of the handwritten samples before training.

A medical billing company needs to extract patient name, date of service, and procedure codes from a variety of referral forms. They have 200 labeled forms. They trained a custom model and achieved 85% accuracy on the test set. They want to improve accuracy before production. What should they do?

Create a composed model with the current model and a layout model.

Switch to the prebuilt health insurance card model.

Publish the model and rely on the confidence score to filter low-confidence results.

Add more labeled samples, especially for forms where extraction failed, and retrain.

Microsoft Certified:Azure AI Engineer Associate

Domain 6Objective 2

Implement an Azure Document Intelligence in Foundry Tools Solution AI-102 Practice Questions (Page 1)

Part of the Implement knowledge mining and information extraction solutions domain, which accounts for 15-20% of the AI-102 exam. Microsoft does not publish an official question count, but from its 100-minute exam (~40–65 total, ~6–13 in this domain), expect 2–4 from this objective — we provide 35 practice questions to prepare you well beyond it. (estimate)

35questions here

7free pages

7concepts

15-20%of the exam

Questions 1–5

1foundation · easy
When you publish a custom document intelligence model, what do you receive that allows applications to call the model?
Select an answer first
2foundation · easy
When you provision an Azure Document Intelligence resource in the Azure portal, which two items are generated that you use to authenticate API calls?
Select an answer first
3foundation · easy
You have a set of 10 sample documents of the same type. You want to train a custom document intelligence model. What is the minimum number of labeled samples you should use for training?
Select an answer first
4application · medium
A manufacturing company has a custom Document Intelligence model that extracts part numbers and quantities from work orders. The model performs well on typed work orders but poorly on handwritten ones. They have a small set of handwritten samples. What should they do to improve performance on handwritten documents?
Select an answer first
5application · medium
A medical billing company needs to extract patient name, date of service, and procedure codes from a variety of referral forms. They have 200 labeled forms. They trained a custom model and achieved 85% accuracy on the test set. They want to improve accuracy before production. What should they do?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to AI-102

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Microsoft. “AI-102” is a trademark of its owner, used for identification only.

Implement an Azure Document Intelligence in Foundry Tools Solution AI-102 Practice Questions (Page 1)

Questions 1–5

When you publish a custom document intelligence model, what do you receive that allows applications to call the model?

When you provision an Azure Document Intelligence resource in the Azure portal, which two items are generated that you use to authenticate API calls?

You have a set of 10 sample documents of the same type. You want to train a custom document intelligence model. What is the minimum number of labeled samples you should use for training?

A medical billing company needs to extract patient name, date of service, and procedure codes from a variety of referral forms. They have 200 labeled forms. They trained a custom model and achieved 85% accuracy on the test set. They want to improve accuracy before production. What should they do?

Keep studying — Implement knowledge mining and information extraction solutions