Amazon
Avaya
Checkpoint
Cisco®
CompTIA
CrowdStrike
EC-Council
EXIN
Fortinet
Google
ISC
ISC2
ITIL®
Juniper
Microsoft
MuleSoft
Okta
Oracle
Palo Alto Networks
Salesforce
All Vendors
  1. Home
  2. Microsoft
  3. AI-900

Microsoft AI-900 Exam (page: 7)
Microsoft Azure AI Fundamentals
Updated on: 29-Aug-2025

Viewing Page 7 of 36
AI-900 PDF 273 Questions

Stating the source of the data used to train a model is an example of which responsible AI principle?

  1. transparency
  2. privacy and security
  3. fairness
  4. reliability and safety

Answer(s): A

Explanation:

Transparency in AI refers to openly sharing information about how an AI system is designed, trained, and operates. Stating the source of the data used to train a model is an example of transparency, as it provides clarity about the origins of the data and helps stakeholders understand the model's development process.



HOTSPOT (Drag and Drop is not supported)
Select the answer that correctly completes the sentence.
Hot Area:

  1. See Explanation section for answer.

Answer(s): A

Explanation:




Box: document Intelligence
A         solution can be used to extract data from scanned invoices.
The Document Intelligence invoice model uses powerful Optical Character Recognition (OCR) capabilities to analyze and extract key fields and line items from sales invoices, utility bills, and purchase orders.


Reference:

https://learn.microsoft.com/en-us/azure/ai-services/document-intelligence/prebuilt/invoice



What can be used to analyze scanned invoices and extract data, such as billing addresses and the total amount due?

  1. Azure AI Custom Vision
  2. Azure AI Search
  3. Azure OpenAI
  4. Azure AI Document Intelligence

Answer(s): D

Explanation:

The Document Intelligence invoice model uses powerful Optical Character Recognition (OCR) capabilities to analyze and extract key fields and line items from sales invoices, utility bills, and purchase orders.


Reference:

https://learn.microsoft.com/en-us/azure/ai-services/document-intelligence/prebuilt/invoice



What should you use to extract details from scanned images of contracts?

  1. Azure AI Document Intelligence
  2. Azure AI Immersive Reader
  3. Azure OpenAI
  4. Azure AI Search

Answer(s): A

Explanation:

Document Intelligence contract model
The Document Intelligence contract model uses powerful Optical Character Recognition (OCR) capabilities to analyze and extract key fields and line items from a select group of important contract entities. Contracts can be of various formats and quality including phone-captured images, scanned documents, and digital PDFs. The API analyzes document text; extracts key information such as Parties, Jurisdictions, Contract ID, and Title; and returns a structured JSON data representation. The model currently supports English-language document formats.


Reference:

https://learn.microsoft.com/en-us/azure/ai-services/document-intelligence/prebuilt/contract



DRAG DROP (Drag and Drop is not supported)
Match the Azure AI service to the appropriate generative AI capability.
To answer, drag the appropriate service from the column on the left to its capability on the right. Each service may be used once, more than once, or not at all.
NOTE: Each correct match is worth one point.
Select and Place:

  1. See Explanation section for answer.

Answer(s): A

Explanation:




Box 1: Azure AI Vision Classify and label images.
To classify and label images, you would typically use Azure AI Vision, which is specifically designed for computer vision tasks. Azure AI Vision can perform image analysis, including classification, object detection, and generating captions. You can also use Azure AI Custom Vision to train custom models that are tailored to your specific needs.
Box 2: Azure OpenAI Service Generate conversational responses
Azure OpenAI Service is designed to be used for generating conversational responses. It utilizes powerful language models like GPT-3.5 Turbo, GPT-4, and GPT-4o, which are specifically optimized for chat completion scenarios, according to Microsoft Learn. These models can handle multi-turn conversations and provide contextually appropriate responses based on user input, effectively simulating conversational AI behavior.
Box 3: Azure AI Speech
Convert speech to text in real time.
The Speech service provides speech-to-text and text-to-speech capabilities with an Azure Speech resource. You can transcribe speech to text with high accuracy, produce natural-sounding text-to-speech voices, translate spoken audio, and use speaker recognition during conversations.
Azure AI Speech service can be used for real-time speech-to-text conversion. It supports both real-time and batch transcription of audio streams into text. Real-time transcription allows for instant transcription of live audio inputs.


Reference:

https://microsoftlearning.github.io/mslearn-ai-vision/Instructions/Exercises/02-image-classification.html https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models https://docs.microsoft.com/en-us/azure/cognitive-services/speech-service/overview



What is an advantage of using a custom model in Document Intelligence?

  1. A custom model can be trained to recognize a variety of form types.
  2. A custom model is less expensive than a prebuilt model.
  3. A custom model always provides higher accuracy.
  4. Only a custom model can be deployed on-premises.

Answer(s): A

Explanation:

In Document Intelligence, a custom model allows you to extract data from specific documents or forms that are unique to your business or industry. This is achieved by training a model on your own labeled data, enabling it to identify and extract information like fields and tables from those documents with high accuracy.
Custom Models vs. Prebuilt Models:
While Document Intelligence offers prebuilt models for common document types like invoices or contracts, custom models allow you to go beyond these general categories and focus on extracting data from your specific forms and documents.
Training with Your Data:
You train a custom model by providing it with labeled samples of your target documents, teaching it to recognize the key fields and layout elements.


Reference:

https://learn.microsoft.com/en-us/azure/ai-services/document-intelligence/train/custom-model



In Azure Machine Learning, what are two ensemble methods for combining models in automated machine learning (automated ML)? Each correct answer presents a complete solution.
NOTE: Each correct selection is worth one point.

  1. voting
  2. computer vision
  3. stacking
  4. classification
  5. regression

Answer(s): A,C

Explanation:

In Azure Machine Learning's automated machine learning (AutoML), two prominent ensemble methods for combining models are Voting and Stacking.


Reference:

https://learn.microsoft.com/en-us/azure/machine-learning/concept-automated-ml 1



Which type of compute resource should you use to attach an existing Azure Kubernetes Service (AKS) cluster to Azure Machine Learning?

  1. compute cluster
  2. serverless compute
  3. inference cluster
  4. compute instance

Answer(s): D

Explanation:

To attach an existing Azure Kubernetes Service (AKS) cluster to Azure Machine Learning, you should use the "Attached Compute" or "Kubernetes Compute Target". This allows you to leverage your existing AKS cluster as a compute resource for your machine learning tasks within Azure Machine Learning. You can achieve this using the Azure CLI v2, Python SDK v2, or Machine Learning Studio UI.
Kubernetes Compute Target:
Azure Machine Learning treats your AKS cluster as a compute target, allowing you to specify it as the location for running your training jobs or deploying models.
2. Attached Compute:
This refers to the ability to connect existing compute resources, like your AKS cluster, to your Azure Machine Learning workspace.


Reference:

https://learn.microsoft.com/en-us/azure/machine-learning/how-to-attach-kubernetes-to-workspace



Viewing Page 7 of 36



Share your comments for Microsoft AI-900 exam with other users:

OLERATO 7/1/2025 5:44:14 AM

great dumps to practice for the exam
SOUTH AFRICA