support@Free-Braindumps.com
Register Login
Cisco
CompTIA
ISC
EC-Council
EXIN
Fortinet
ITIL
Juniper
Microsoft
VMware
Avaya
SAP
Google
NVIDIA
Salesforce
Amazon
Palo Alto Networks
ISC2
Splunk
ServiceNow
All Vendors
  1. Home
  2. Exams
  3. Microsoft
  4. AI-901

Microsoft AI-901 Exam Actual Questions
Microsoft Azure AI Fundamentals (Page 3 )

Updated On: 23-Jul-2026
Page 2 of 8
PDF with 52 Questions

You are using the Azure Speech SDK to develop a Python application that supports real-time spoken conversations.
Which Azure Speech class should you use to configure the connection to the Azure Speech service?

  1. AudioOutputConfig
  2. SpeechSynthesizer
  3. AudioConfig
  4. SpeechConfig

Answer(s): D

Explanation:

To configure the connection to the Azure Speech service, you should use the azure.cognitiveservices.speech.SpeechConfig class.
Implementation Details Initialization: This class holds authentication and endpoint details, such as your Azure resource key and region.
Usage: You pass this configuration object into your real-time processing classes, like SpeechRecognizer (for speech-to-text) or SpeechSynthesizer (for text-to-speech).


Reference:

https://docs.azure.cn/en-us/ai-services/speech-service/how-to-recognize-speech



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: a base64-encoded image data When using the OpenAI Responses API and a vision-enabled model, you can include an image in a request by providing the image as _______________.
When using the OpenAI Responses API with a vision-enabled model, you can include an image in a request by providing it in three different formats: A fully qualified image URL: Passing the web link where the image is hosted online. *-> A Base64-encoded data URL: Encoding a local image file directly into a text string (formatted as data:image/jpeg;base64,...). A file ID: Uploading the image via the Files API first and referencing its unique ID.


Reference:

https://developers.openai.com/api/docs/guides/image-generation



You need to create an AI agent in Microsoft Foundry that follows a specific role and behavior when responding to users.
What should you configure?

  1. system instructions
  2. temperature
  3. tokens per minute (TPM)
  4. max completion tokens

Answer(s): A

Explanation:

In Microsoft Foundry, instructions (formerly referred to as system messages) are the primary way to define the role, behavior, tone, and constraints of your AI agent.


Reference:

https://learn.microsoft.com/en-us/azure/foundry-classic/agents/quickstart



HOTSPOT (Drag and Drop is not supported)
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
Note: Each correct selection is worth one point.
Hot Area:

  1. See Explanation section for answer.

Answer(s): A

Explanation:



Box 1: No No - System prompts can be used to authorize users.
In Azure AI, system prompts cannot securely authorize users. They guide model behavior, but because LLMs are susceptible to prompt injections and jailbreaks, relying on them for security poses severe vulnerabilities.Instead, you should handle authorization and authentication through secure infrastructural controls and middleware.
Box 2: Yes Yes - A system prompt used to reduce tokens per minute (TPM).
A system prompt can be specifically engineered to significantly reduce your Tokens Per Minute (TPM) consumption in Azure OpenAI Service. TPM limits are triggered by the combined volume of input (prompt) tokens and output (completion) tokens.
The primary mechanism of a "TPM-reduction system prompt" is to strictly constrain the model’s output verbosity and eliminate conversational filler, which prevents your application from hitting 429 TooManyRequests throttling errors.
Box 3: Yes Yes - A system prompt guides the behavior of a generate AI model.
In Azure AI (and generative AI in general), a system prompt (often called a system message or metaprompt) acts as the foundational instruction set that dictates the AI model's behavior, tone, operational constraints, and guardrails.A system prompt serves as a "foundational job description" for the model. It sits in the backgro


Reference:

https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/system-message



You are developing a web app that processes invoices to calculate expenses.
You need to extract structured fields, including nested values, from the invoices by using a defined schema.
What should you use?

  1. an optical character recognition (OCR)-only document processing pipeline
  2. a transcription workflow in Azure Speech in Foundry Tools
  3. an analyzer in Azure Content Understanding in Foundry Tools
  4. an Azure AI Search service

Answer(s): C

Explanation:

An analyzer in Azure Content Understanding within Foundry Tools can be used for this scenario.
Prebuilt Invoice Support: It offers prebuilt invoice analyzer templates specifically trained to identify common transactional data out of the box.
Schema-Driven Extraction: Instead of manual prompt engineering or tedious labeling, you can supply a strict JSON schema containing the specific fields you need.
Nested Values & Layouts: It handles hierarchical, nested elements (such as itemized tables, nested line items, or customer details) and maps them cleanly into arrays or structured objects.
Web App Integration: Activating the analyzer creates a highly scalable REST API endpoint. Your invoice-processing web app can query this endpoint to receive clean, machine-readable structured JSON payloads ready for downstream expense calculation.


Reference:

https://learn.microsoft.com/en-us/azure/ai-services/content-understanding/overview



HOTSPOT (Drag and Drop is not supported)
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
Note: Each correct selection is worth one point.
Hot Area:

  1. See Explanation section for answer.

Answer(s): A

Explanation:



Box 1: No No - Voice Live returns only transcribed text.
Azure AI's Voice Live API does not return only transcribed text. It is a fully managed, end-to-end, low-latency speech-to-speech pipeline.Instead of forcing you to build separate steps for speech-to-text, LLM processing, and text-to-speech, the Azure Voice Live API combines these into a single WebSocket connection.
Box 2: No No - Voice Live requires you to separately implement speech to text and text to speech services.
You do not need to implement them separately. The Azure Voice Live API is a fully managed service that integrates speech-to-text, generative AI, and text-to-speech into a single unified pipeline. It handles both listening and speaking through one WebSocket interface, eliminating complex patchwork code.
Box 3: Yes Voice Live combines speech to text, reasoning, and text to speech into a single conversational experience.
The Azure Voice Live API is a fully managed service that integrates speech-to-text, generative AI, and text-to-speech into a single unified pipeline.


Reference:

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/voice-live-language-support



HOTSPOT (Drag and Drop is not supported)
You are developing a voice application that listens for spoken commands and converts them into text by using Azure Speech in Foundry Tools.
How should you complete the Python code? To answer, select the appropriate option in the answer area.
Note: Each correct selection is worth one point.
Hot Area:

  1. See Explanation section for answer.

Answer(s): A

Explanation:





Box: recognizer.recognize_once() The recognizer.recognize_once() method belongs to the SpeechRecognizer object. Its purpose is to start speech recognition from your audio input (like a microphone) and stop after a single utterance is parsed into text. This perfectly fulfills the goal of creating an application that "listens for spoken commands and converts them into text".
Purpose: The recognizer.recognize_once() or recognizer.recognize_once_async() method instructs the Azure Speech Recognizer to start listening to the audio input stream.
Behavior: It captures a single spoken utterance (like a short command), performs speech recognition, and returns the recognized text.


Reference:

https://gotranscript.com/public/developing-speech-to-text-with-c-sharp-sdk



What are two purposes of instructions when prompting a generative AI model? Each correct answer presents part of the solution.
Note: Each correct selection is worth one point.

  1. defines the agent’s role and behavior
  2. selects which model to use
  3. defines the Azure region where inference occurs
  4. defines the tokens per minute (TPM) allocation for the model
  5. defines constraints on the model’s responses

Answer(s): A,E

Explanation:

Giving a generative AI model clear, structural instructions (commonly referred to as Prompt Engineering) is the standard method used to define its persona, guide its behavior, and place strict operational boundaries on its responses.This is typically achieved through explicit instruction sets like System Instructions and targeted Prompt Guidelines.
1. Defining the Agent's Role and BehaviorEstablishing a persona prevents the AI from wandering into areas outside its intended scope and ensures the delivery style matches your needs.
2. Defining Constraints on ResponsesConstraints act as guardrails that prevent the AI from hallucinating, rambling, or providing incorrect information.


Reference:

https://ai.google.dev/gemini-api/docs/prompting-strategies



  PDF
Viewing page 3 of 8
Viewing questions 17 - 24 out of 52 questions


Post your Comments and Discuss Microsoft AI-901 exam prep with other Community members:

AI Tutor AI Tutor 👋 I’m here to help!