Microsoft AI-102 Designing and Implementing a Microsoft Azure AI Solution Exam Practice Test

To detect aggressive language (e.g., profanity/abuse) and sexually explicit content in text using the Content Moderator Text – Screen API:

classify = true

Enables machine-assisted text classification that returns category scores (e.g., sexually explicit/suggestive categories). This is required to identify sexual content.

language = < code > (for example, eng )

Specifies the language so profanity/abusive terms are detected correctly. The profanity detection in Text-Screen is language dependent.

autocorrect = true

Helps normalize intentionally misspelled or obfuscated words so that profanity or aggressive terms are still detected (e.g., “f@#$” → “fuck”).

Other parameters in the blade are not necessary for this goal:

PII targets personal data extraction, not aggression/sexuality.

listId is for matching against custom blocklists.

Headers like Ocp-Apim-Subscription-Key and Content-Type are required to call the API, but they are not the content-analysis settings that control detection behavior asked for in this question.

Microsoft Azure AI References

Azure AI Content Moderator – Screen text API parameters (autocorrect, language, classify) and category outputs.

Azure AI Content Moderator – Text moderation concepts: profanity detection and classification categories.

The requirement: A chatbot that answers based on a product knowledgebase (multiple PDFs) .

Best choice: Custom Question Answering in Azure AI Language .

It allows you to ingest documents like PDFs, FAQs, manuals.

Minimal development effort since you just upload documents and the service automatically indexes them.

Other options:

CLU → for intent/intent-based classification, not for knowledgebases.

Language detection → only detects language, not Q & A.

Azure OpenAI → could work, but more costly and requires prompt engineering (not the “minimize effort/cost” requirement).

Microsoft Reference:

Custom question answering in Azure AI Language

Comprehensive Detailed Explanation

You’re adding multiple images for each person into a PersonGroup using the Face API SDK.

File.OpenRead(imagePath) returns a Stream, so the using declaration must declare Stream t = File.OpenRead(imagePath) .

To upload an image from a stream to a person in a person group, call

faceClient.PersonGroupPerson.AddFaceFromStreamAsync(personGroupId, personId, t);

Putting it together:

Parallel.For( 0 , PersonCount, async i = >

{

Guid personId = persons[i].PersonId;

string personImageDir = $ " /path/to/person/ {i} /images " ;

foreach ( string imagePath in Directory.GetFiles(personImageDir, " *.jpg " ))

{

using (Stream t = File.OpenRead(imagePath))

{

await faceClient.PersonGroupPerson.AddFaceFromStreamAsync(

personGroupId, personId, t);

}

}

});

This loops the images for each person, opens each as a stream, and adds the face to the specified person in the person group.

References

Microsoft Docs – Face API: Add a face to a person ( AddFaceFromStreamAsync , AddFaceFromUrlAsync ) in PersonGroupPerson operations.

https://learn.microsoft.com/azure/ai-services/face/ (see SDK/PersonGroupPerson methods)

NET File.OpenRead returns Stream .

https://learn.microsoft.com/dotnet/api/system.io.file.openread

You want to prevent users from sharing sexually explicit images and minimize development effort .

Azure AI Content Safety Studio provides a no-code, browser-based experience to test and tune image moderation using the Content Safety service. It supports out-of-the-box classification for Sexual , Hate , Violence , and Self-harm categories, lets you adjust severity thresholds, review results, and export settings or sample code with minimal engineering work.

Other options:

A. Visual Studio – an IDE, not a moderation service.

B. Vision Studio in Azure AI Vision – focuses on vision tasks (OCR, image analysis). For policy/compliance moderation of explicit content, Content Safety is the correct service.

D. Azure AI Studio – general hub for building generative AI apps; not the specialized moderation experience.

Microsoft References

Azure AI Content Safety overview and image moderation concepts.

Content Safety Studio quickstart and scenarios.

Content categories (Sexual, Hate, Violence, Self-harm).

The agent will reason over the uploaded file.●○

The code will create a run and check the output.○●

The code will create an agent and enable a file search.○●

 The agent will reason over the uploaded file.

Yes : The code includes file = project_client.agents.upload_file_and_poll(file_path= " ./data/file1.zip " , purpose=FilePurpose.AGENTS). The upload_file_and_poll method with FilePurpose.AGENTS suggests that the file (file1.zip) is uploaded for the agent to process, and agents in the Azure AI Foundry Agent Service are typically designed to reason over uploaded data. This indicates the agent will reason over the uploaded file.

 The code will create a run and check the output.

No : The provided code snippet does not include any explicit call to create a run (e.g., a method like create_run or start_run) or check the output (e.g., a method to retrieve or poll run results). The code only uploads a file and creates a vector store, but it does not initiate or monitor a run.

 The code will create an agent and enable a file search.

No : The code does not contain a method to create an agent (e.g., create_agent) or explicitly enable a file search. The create_vector_store_and_poll method creates a vector store, which is typically used for indexing data for search, but there is no indication of enabling a file search or creating an agent in this snippet.

Step 1: For the knowledge base, select Show active learning suggestions.

In order to see the suggested questions, on the Edit knowledge base page, select View Options, then select Show active learning suggestions.

Step 2: Approve and reject suggestions.

Each QnA pair suggests the new question alternatives with a check mark, ✔ , to accept the question or an x to reject the suggestions. Select the check mark to add the question.

Step 3: Save and train the knowledge base.

Select Save and Train to save the changes to the knowledge base.

Step 4: Publish the knowledge base.

Select Publish to allow the changes to be available from the GenerateAnswer API.

When 5 or more similar queries are clustered, every 30 minutes, QnA Maker suggests the alternate questions for you to accept or reject.

You are asked to build an app that translates speech from English into French, German, and Spanish using Azure AI Speech Translation.

Step 1 – Define target languages

The SpeechTranslationConfig object requires language codes (not full names like " French " , " German " , " Spanish " ).

Valid language codes:

French → " fr "

German → " de "

Spanish → " es "

Therefore, the correct syntax for the output languages is:

var languages = new List < string > { " fr " , " de " , " es " };

Step 2 – Choose the recognizer type

IntentRecognizer : Used for intent recognition with LUIS, not translation.

SpeakerRecognizer : Used for speaker identification/verification, not translation.

SpeechSynthesizer : Used for text-to-speech, not translation.

TranslationRecognizer : Used specifically for speech-to-speech or speech-to-text translation.

Therefore, the correct recognizer is TranslationRecognizer.

Correct Answer:

Output Languages: { " fr " , " de " , " es " }

Recognizer: TranslationRecognizer

Microsoft References

Speech Translation with Azure AI Speech

TranslationRecognizer Class (C# API)

Supported language codes

Azure AI Video Indexer (VI) imposes limits on both file duration and file size:

Duration: Up to 6 hours for all presets; Basic Audio preset supports up to 12 hours.

Size: When uploading from your device, the limit is 2 GB (uploading by URL allows up to 30 GB).

Formats: MP4, AVI (video) and MP3 (audio codec/file) are supported.

Evaluating each file:

File1 – MP4, 34 mins, 1,500 MB: Within 6-hour duration and under 2 GB → Supported.

File2 – AVI, 500 mins (≈8h20m), 1,700 MB: Exceeds the 6-hour limit → Not supported.

File3 – MP3, 300 mins (5h), 980 MB: Under 6 hours (and also under the 12-hour Basic Audio limit) and under 2 GB → Supported.

File4 – MP4, 350 mins (≈5h50m), 2,800 MB: Duration is under 6 hours but size exceeds 2 GB for device uploads → Not supported (unless uploaded by URL, which the question doesn’t state).

Therefore, the files you can upload and analyze (under standard device upload constraints) are File1 and File3 only.

Microsoft References

Support matrix & limits — file duration (6h for all presets; 12h for Basic Audio), device upload size limit 2 GB; supported formats (MP4, AVI, WAV, etc.) and audio codecs including MP3. Microsoft Learn

Release notes — increase of duration limit from 4h to 6h (and 12h for Basic Audio). Microsoft Learn

Upload guidance — device upload 2 GB vs 30 GB via URL.

Comprehensive Detailed Explanation

Scenario:

You have a chatbot using Question Answering in Azure Cognitive Service for Language.

Users report that the chatbot’s responses to spurious questions (nonsense or off-topic input) lack formality.

The goal is to ensure the chatbot provides formal responses.

Proposed Solution:

Change the chitchat source to qna_chitchat_friendly.tsv in Language Studio, retrain, and republish.

Analysis:

Azure provides prebuilt chitchat datasets for handling spurious/off-topic inputs.

These datasets come in three styles:

Professional → formal tone

Friendly → casual/informal tone

Witty → humorous tone

If you want formal responses, you must use the professional chitchat dataset ( qna_chitchat_professional.tsv ).

The proposed solution uses qna_chitchat_friendly.tsv , which produces informal/casual responses, not formal ones.

Therefore, the solution does not meet the goal.

Correct Answer: B. No

Microsoft References

Add chit-chat to your QnA Maker knowledge base

Question Answering with chit-chat styles

Removing all the chit-chat question and answer pairs from the project will not ensure that the chatbot provides formal responses to spurious questions. It will only make the chatbot unable to handle any chit-chat scenarios, which may result in a poor user experience and a loss of engagement. Instead, you should choose a chit-chat personality that matches the tone and style of your chatbot, such as Professional or Caring. You can also edit the chit-chat questions and answers to suit your specific needs, or add new ones that are not in the predefined data set12. This way, you can ensure that the chatbot responds appropriately to spurious questions, while still maintaining a conversational and engaging interaction with the user.

The requirement is to ensure the chatbot provides formal responses to spurious questions (random/off-topic).

The correct way to handle this is to add a chit-chat personality dataset (such as qna_chitchat_professional.tsv ) in Language Studio, retrain, and republish.

Modifying existing QnA pairs for custom intents (as the solution suggests) would not cover spurious/off-topic questions automatically. It only affects known knowledge base intents.

Therefore, this solution does not meet the goal.

Correct Answer: B

B: The Text Analytics API ' s Sentiment Analysis feature provides two ways for detecting positive and negative sentiment. If you send a Sentiment Analysis request, the API will return sentiment labels (such as " negative " , " neutral " and " positive " ) and confidence scores at the sentence and document-level.

D: The Language Detection feature of the Azure Text Analytics REST API evaluates text input for each document and returns language identifiers with a score that indicates the strength of the analysis.

This capability is useful for content stores that collect arbitrary text, where language is unknown. Reference:

https://docs.microsoft.com/en-us/azure/cognitive-services/text-analytics/how-tos/text-analytics-how-to- sentiment-analysis?tabs=version-3-1

https://docs.microsoft.com/en-us/azure/cognitive-services/text-analytics/how-tos/text-analytics-how-to- language-detection

You are developing an app that uses the Azure AI Speech service , and you must ensure the app can authenticate using a Microsoft Entra ID (Azure AD) token , rather than the standard key-based authentication.

According to the official Microsoft documentation :

“To authenticate with a Microsoft Entra token, the Speech resource must have a custom subdomain and use a private endpoint. The Speech service uses custom subdomains with private endpoints only.”

— Microsoft Learn: Azure AI Speech service - Role-based access control

Therefore, two prerequisites must be configured for Entra ID token-based authentication:

1. Create (Certify) a Custom Subdomain (Option D)

When you create an Azure AI Speech resource, you can specify a custom subdomain name (e.g., contoso.cognitiveservices.azure.com).

The custom subdomain ensures that each Speech endpoint has a unique URL.

Microsoft Entra token authentication depends on this subdomain for token validation.

Without a custom subdomain, Entra ID tokens cannot be used for authentication.

Hence, this is required.

2. Create a Private Endpoint (Option B)

For Microsoft Entra ID authentication, the Speech resource must be secured within a private endpoint .

Private endpoints provide a secure, private connection from your virtual network to the Azure AI Speech resource.

The Speech service only supports Entra ID token authentication when the resource is configured with a private endpoint.

Hence, this is required.

Why Not the Other Options:

A. Create a Conditional Access – Conditional Access policies restrict user access but do not configure token authentication for services.

C. Request an X.509 certificate – Certificates are not used for Speech service authentication.

E. Enable a virtual network service endpoint – The Speech service uses private endpoints , not service endpoints, for Entra ID token authentication.

Dimensional models and data warehouses are designed for OLAP workloads (aggregations, analytics, reporting).

OLTP is for transaction-based line-of-business applications.

Batch processing is about moving and transforming data, not querying.

Stream processing handles continuous real-time data, not historical warehouse queries.

Correct Answer: B

You are building a chatbot that will use an Azure OpenAI model.

You need to deploy the model in your Azure subscription.

Key Steps to Deploy an Azure OpenAI Model for a Chatbot:

Apply for access to Azure OpenAI

Azure OpenAI is a gated service and requires approval before you can provision resources.

Provision an Azure OpenAI resource

After approval, you must provision an Azure OpenAI resource in your subscription.

Deploy the GPT model

Since the chatbot requires natural language understanding and conversation, you deploy a GPT model (e.g., GPT-4, GPT-3.5).

Actions That Are Not Required for this Question:

Provision Azure API Management → Not needed for basic deployment (used only for advanced API management scenarios).

Deploy the embeddings model → Used for search/RAG, not necessary for chatbot baseline.

Deploy the DALL-E model → Used for image generation, not chatbot text responses.

Correct Order:

Apply for access to Azure OpenAI.

Provision an Azure OpenAI resource.

Deploy the GPT model.

Verified Answer (Sequence):

Apply for access to Azure OpenAI

Provision an Azure OpenAI resource

Deploy the GPT model

Microsoft References:

Quickstart: Get started with Azure OpenAI

Deploying models in Azure OpenAI

Azure Metrics Advisor builds on Anomaly Detector but adds root cause analysis , correlation across multiple signals, incident detection, and alerting out-of-the-box.

D. Anomaly Detector → Works for single-signal anomaly detection but lacks correlation and root cause analysis.

C. Azure Machine Learning → Would work but requires custom model development (more dev effort).

B. Form Recognizer → Used for document processing, not sensor telemetry.

Correct Answer: A

When working with Speech Studio in Azure AI Speech (for example, to build a custom speech model for recognition or training purposes), you must provide speech samples along with their transcripts.

Requirements for Uploading Speech Samples:

Audio File Format

The supported format is .wav (PCM, 16-bit, 16 kHz or 8 kHz mono).

Other formats like .wma , .flac , or .mp3 are not accepted for training custom models.

Batch Upload via .zip

Audio files must be placed in a .zip archive.

The .zip file must also include a transcription file (usually a .txt or .json file) that maps the spoken audio in each .wav file to the correct text.

Transcript Format

Transcripts must be plain text, not Microsoft Word documents.

The transcript must align correctly with each audio file to ensure the Speech service learns from accurate mappings.

Option Analysis

A. Combine into one .wma file

Incorrect. .wma is not supported. Also, Speech Studio expects multiple .wav samples, not one big file.

B. Upload a .zip file containing .wav files + transcript

Correct. This is the officially supported method for uploading training data into a Speech Studio project.

C. Upload .flac files with Word transcript

Incorrect. .flac is not supported. Transcripts must be plain text, not Word files.

D. Upload individual .wma files

Incorrect. .wma is not supported.

Correct Answer: B. Upload a .zip file that contains a collection of audio files in the .wav format and a corresponding text transcript file.

Microsoft References

Prepare training data for Custom Speech

Speech Studio – Data Requirements

Event log data is typically semi-structured (key-value pairs, JSON-like format).

Azure Table Storage is a NoSQL key-value store that is ideal for storing semi-structured, non-relational datasets .

Logs arrive as events occur, so they can be appended as rows in table storage.

Other options:

Azure Queue storage → Used for messaging between application components, not for storing logs.

Azure Files → Cloud-based SMB file shares, not optimized for event log storage.

Correct Answer: Azure Table storage

To minimize development effort in Azure AI Content Understanding, you should start from a video analyzer template that already targets large libraries and general-purpose metadata extraction. The Media asset management template is designed exactly for this scenario—organizing and extracting structured information from large collections of videos (news, marketing, broadcast, TV episodes, film archives). Templates are fully customizable, so you can add a field such as sentiment per video or per segment without building the schema from scratch. This approach yields the required sentiment analysis while keeping effort low.

“Video shot analysis” is not a Content Understanding template (shot detection is a capability inside the video analyzer), and the Advertising template is tuned for ad-specific fields (brand presence, ad categories), which is unnecessary for a generic video library. Using Media asset management aligns with the library use case and lets you add a sentiment field to the template’s schema.

Microsoft References

Analyzer templates (Video) — lists the available video templates including Media asset management and Advertising and notes that templates are customizable. Microsoft Learn

Video solutions overview — explains prebuilt video analyzer, shot detection (as a capability), and shows how to customize fields (example includes extracting fields such as sentiment per scene ). Microsoft Learn

Conversational Language Understanding (CLU) is trained using utterances .

Adding these example phrases as utterances for the FindContact intent is exactly how training is done.

This ensures the model learns to recognize similar queries.

Correct Answer: Yes

Relational databases use SQL (Structured Query Language) to query and manipulate data.

B is incorrect: Relationships exist between tables (foreign keys, joins).

C is incorrect: Flexible schema is a characteristic of NoSQL, not relational.

D is incorrect: Relational databases aim to reduce duplicate data through normalization.

The requirement is to analyze video content to identify mentions of specific company names. This is essentially an audio transcript analysis task (speech-to-text followed by entity recognition). Azure Video Analyzer for Media integrates with Content model customization to detect such entities.

Step-by-step reasoning:

Sign in to the Azure Video Analyzer for Media website.

Azure Video Analyzer for Media (formerly Video Indexer) is the service used for analyzing videos and extracting insights such as speech, text, brands, and entities.

From Content model customization, select Language.

To recognize specific company names in transcripts, you must customize the Language model.

Customization allows you to add vocabulary (e.g., company names) so that speech recognition and transcription are tuned to correctly identify them.

Add the specific company names to the include list.

Adding them to the include list ensures the system actively looks for and recognizes those company names during transcript analysis.

The exclude list would suppress terms, not detect them, so it is not correct in this scenario.

Why not other options?

Sign in to the Custom Vision website → Custom Vision is for image classification/object detection, not audio/video transcript entity recognition.

From Content model customization, select Brands → The Brands model helps detect logos and existing brand mentions, but for custom company names not in the brand catalog, you must customize the Language model.

Exclude list → Would filter out terms, which is opposite of the requirement.

Correct Answer Order:

Sign in to the Azure Video Analyzer for Media website.

From Content model customization, select Language.

Add the specific company names to the include list.

Microsoft References

Customize models in Video Indexer

Azure Video Analyzer for Media overview

Content model customization in Video Indexer

You are creating an SSML file for a text-to-speech app using a custom neural voice.

Requirements:

Expresses a calm tone → corresponds to style (e.g., " gentle " ).

Imitates the voice of a young adult female → corresponds to role (e.g., " YoungAdultFemale " ).

Key Knowledge:

In Speech Synthesis Markup Language (SSML) with Microsoft Azure TTS:

< mstts:express-as > supports attributes such as style, role, styledegree, voice, etc.

role → controls persona/age/gender (e.g., " YoungAdultFemale " ).

style → controls the speaking tone/intonation (e.g., " gentle " , " calm " , " chat " ).

Code Completion:

< mstts:express-as role = " YoungAdultFemale " style = " gentle " >

How can I assist you?

< / mstts:express-as >

Verified Answer:

First dropdown: role

Second dropdown: style

Microsoft References:

Azure Speech SSML – express-as element

Speech synthesis styles and roles

✅ Final Answer:

role = " YoungAdultFemale "

style = " gentle "

Project type: Object detection

Domain: General (compact)

The scenario requires building a Custom Vision model for a factory with intermittent internet connectivity. The solution must:

Detect defects in packaging

Locate the defects (provide bounding box information)

Ensure each package contains four products

Support offline use due to intermittent connectivity

Step 1 – Determine Project Type

Image Classification: Assigns a label to an image, but cannot locate where objects/defects are in the image.

Object Detection: Identifies what objects are present and their location (via bounding boxes).

Since the requirement is to identify location of defects and count products, the correct choice is Object detection.

Step 2 – Determine Domain

General domain: Standard general-purpose domain, optimized for cloud inference.

General (compact) domain: A lightweight version designed for exporting models to run offline on edge devices (iOS, Android, ONNX, etc.). Perfect for scenarios with intermittent internet connectivity.

Food domain: Specialized for food image classification, not defect detection in packaging.

Logo domain: Specialized for logo recognition, irrelevant here.

Since the requirement includes offline capability, the correct domain is General (compact).

Correct Answer:

Project type: Object detection

Domain: General (compact)

Microsoft References

Custom Vision project types: Classification vs Object detection

Domains in Custom Vision (General, General compact, Food, Logo, etc.)

Exporting compact models for offline use

To analyze receipts from a URL, use the StartRecognizeReceiptsFromUri method

Example code:

private static async Task AnalyzeReceipt(

FormRecognizerClient recognizerClient, string receiptUri)

{

RecognizedFormCollection receipts = await recognizerClient.StartRecognizeReceiptsFromUri(new Uri(receiptUrl)).WaitForCompletionAsync();

Your app uses the primary admin key for index management. If it’s suspected compromised, the least-downtime rotation is:

Regenerate the secondary admin key to ensure it’s known only to you.

Switch the app to use the secondary key (no downtime—primary still works until you cut it off).

Regenerate the primary key to invalidate the compromised credential.

Query keys are read-only and cannot manage indexes, so switching to a query key would not meet the requirement. Regenerating the primary first would break the app until you update it.

References (Microsoft Azure Cognitive Search)

Admin (primary/secondary) vs. query keys and key rotation; key regeneration invalidates only the selected key.

Guidance to use the alternate admin key while rotating the other to avoid downtime.

You need to prepare training data for Azure OpenAI custom model fine-tuning using the OpenAI CLI data preparation tool .

Supported input formats: JSONL (primary).

CLI pre-processing accepted formats: TSV, CSV, JSON.

Not supported: PDF, XML, XLSX directly.

Checking files:

File1.tsv (80 MB) → Supported (CLI can convert TSV to JSONL).

File2.xml (25 MB) → Not supported (XML not accepted).

File3.pdf (50 MB) → Not supported.

File4.xlsx (200 MB) → Not supported directly.

Correct Answer: A. File1.tsv only

Azure Synapse Analytics (formerly Azure SQL Data Warehouse) uses a massively parallel processing (MPP) architecture for big data and analytics workloads.

Control Node:

Acts as the brain of the system.

Receives queries, optimizes them, and then coordinates execution.

Does not perform large-scale data crunching itself.

Compute Nodes:

Perform the actual parallel execution of queries.

Data is distributed across these nodes.

The MPP engine breaks down queries into smaller tasks and distributes them to compute nodes for fast, parallel execution.

Review of Answer Options:

Distributes processing across compute nodes → Correct. This is the essence of MPP.

Distributes processing across control nodes → Incorrect. The control node does not process queries in parallel; it coordinates them.

Redirects client connections across compute nodes → Incorrect. Client connections go to the control node, not directly to compute nodes.

Redirects client connections across control nodes → Incorrect. There is typically one control node per Synapse instance.

Correct Answer: Distributes processing across compute nodes

Microsoft References

Azure Synapse Analytics architecture

Distributed query processing in Azure Synapse

You’re using Azure AI Language with sensitive data and need to ensure only specific Azure processes (subnets) can access the resource with minimal administration .

Configure Virtual Network (VNet) rules on the Azure AI (Cognitive Services) resource and enable service endpoints (Microsoft.CognitiveServices) on the selected subnets. This restricts access to requests originating from those VNets/subnets—no need to deploy or manage gateways or IPsec tunnels, keeping operational overhead low. If even stronger isolation is needed later, you can move to private endpoints , but VNet rules + service endpoints meet the requirement and minimize admin effort.

Other options:

A. Azure Application Gateway – L7 load balancer; does not restrict service access by itself.

B. IPsec rules – would require VPN/ExpressRoute setup and more admin overhead.

C. Virtual network gateway – used for cross-premises connectivity, not needed just to restrict access to the Language service.

Microsoft References

Azure AI services network isolation: virtual networks, firewall, and service endpoints.

Configure virtual network rules for Cognitive Services (Microsoft.CognitiveServices service endpoint).

Private endpoints for Azure AI services (for future stronger isolation, if needed).

 Stream processing has access to the most recent data received or data within a rolling time window. → Yes

 Batch processing must occur immediately and have latency in the order of seconds or milliseconds. → No

 Stream processing is used for simple response functions, aggregates, or calculations such as rolling averages. → Yes

Comprehensive Detailed Explanation

Stream processing has access to the most recent data received or data within a rolling time window.

Correct.

Stream processing handles real-time or near real-time data.

It operates on a continuous data stream and often uses a rolling time window for analytics.

Answer: Yes

Batch processing must occur immediately and have latency in the order of seconds or milliseconds.

Incorrect.

Batch processing is designed for large volumes of data collected over time.

It typically has high latency (minutes, hours, or even days), not immediate execution.

Answer: No

Stream processing is used for simple response functions, aggregates, or calculations such as rolling averages.

Correct.

Common stream analytics tasks include event-driven responses, real-time aggregations, anomaly detection, and rolling averages.

Answer: Yes

Correct Answers:

Yes

No

Yes

Microsoft References

Batch vs. Stream processing in Azure

Azure Stream Analytics overview

- enable active learning

- validate the utterances

- train and republish

https://docs.microsoft.com/en-us/azure/cognitive-services/luis/luis-how-to-review-endpoint-utterances#log-user-queries-to-enable-active-learning

https://docs.microsoft.com/en-us/azure/cognitive-services/luis/luis-concept-prebuilt-model

The requirement is:

" Measure the public perception of your brand on social media using natural language processing (NLP). "

A. Content Moderator

Designed to filter offensive or inappropriate content in text, images, and videos.

Not suitable for sentiment analysis or measuring perception.

B. Form Recognizer

Extracts structured data from documents (invoices, receipts, forms).

Not used for analyzing social media posts.

C. Computer Vision

Analyzes images and videos for object detection, image classification, OCR, etc.

Not used for text-based sentiment analysis.

D. Language service ✅

Provides natural language processing (NLP) capabilities such as sentiment analysis, opinion mining, key phrase extraction, entity recognition , and language detection .

Perfect fit for analyzing public perception from social media posts .

Step 1: The Authorization Header

When calling Azure AI Document Intelligence REST API, the correct header to authenticate is:

-H " Ocp-Apim-Subscription-Key: {yourKey} "

Key1 and Secret are not valid header names.

Subscription-Key is also not the standard header.

Correct choice: Ocp-Apim-Subscription-Key

Step 2: Selecting the Prebuilt Model

Since the file contains tabular data (tables, structured document), the right prebuilt model is:

prebuilt-document → extracts text, tables, structure, key-value pairs.

prebuilt-contract → specialized for contracts.

prebuilt-layout → extracts layout, lines, and words but not tables in structured form.

prebuilt-read → simple OCR only.

So the correct choice is: prebuilt-document

Final Completed Command

curl -i -X POST " {endpoint}/formrecognizer/documentModels/prebuilt-document:analyze?api-version=2023-07-31 " \

-H " Content-Type: application/json " \

-H " Ocp-Apim-Subscription-Key: {yourKey} " \

-d " { ' urlSource ' : ' https:// < your-storage > /test.pdf ' } "

Verified Answer:

Left box: Ocp-Apim-Subscription-Key

Right box: prebuilt-document

Microsoft References:

Azure AI Document Intelligence REST API

Prebuilt Document model

PaaS databases reduce administrative overhead because Microsoft manages backups, patching, updates, and infrastructure.

Users always get access to the latest features without manual upgrades.

B is incorrect: Backup/restore is automated; you don’t get full manual control.

C is incorrect: In-database ML services are supported in SQL Server (on-prem / IaaS), but not natively in Azure SQL Database.

To access multiple Azure AI capabilities (e.g., Speech and Language) using a single endpoint and a single set of credentials, you should provision an Azure AI multi-service resource (called Azure AI service in the portal; formerly “Cognitive Services” multi-service account). This resource provides one endpoint and one key (or Microsoft Entra ID) that authorizes calls across supported services, including Speech and Language. Creating separate “Azure AI Speech” or “Azure AI Language” resources would give you distinct endpoints/keys for each, which does not meet the requirement. Azure AI Content Safety is unrelated.

Microsoft References

Azure AI services multi-service resource (single endpoint/credential across services).

https://learn.microsoft.com/azure/ai-services/multi-service-resource

What are Azure AI services (overview of consolidated access model).

https://learn.microsoft.com/azure/ai-services/what-are-ai-services

Box 1: Azure Blob storage

At the start of the pipeline, you have unstructured text or non-text content (such as images, scanned documents, or JPEG files). Data must exist in an Azure data storage service that can be accessed by an indexer.

Box 2: Computer Vision API

Scenario: Provide users with the ability to search insight gained from the images, manuals, and videos associated with the products.

The Computer Vision Read API is Azure ' s latest OCR technology (learn what ' s new) that extracts printed text (in several languages), handwritten text (English only), digits, and currency symbols from images and multi-page PDF documents.

Box 3: Translator API

Scenario: Product descriptions, transcripts, and all text must be available in English, Spanish, and Portuguese.

Box 4: Azure Files

Scenario: Store all raw insight data that was generated, so the data can be processed later.

Scenario: Support autocompletion and autosuggestion based on all product name variants.

A: Call a suggester-enabled query, in the form of a Suggestion request or Autocomplete request, using an API. API usage is illustrated in the following call to the Autocomplete REST API.

POST /indexes/myxboxgames/docs/autocomplete?search & api-version=2020-06-30

{

" search " : " minecraf " ,

" suggesterName " : " sg "

}

B: In Azure Cognitive Search, typeahead or " search-as-you-type " is enabled through a suggester. A suggester provides a list of fields that undergo additional tokenization, generating prefix sequences to support matches on partial terms. For example, a suggester that includes a City field with a value for " Seattle " will have prefix combinations of " sea " , " seat " , " seatt " , and " seattl " to support typeahead.

F. Use the default standard Lucene analyzer ( " analyzer " : null) or a language analyzer (for example, " analyzer " : " en.Microsoft " ) on the field.

Scenario: All videos must have transcripts that are associated to the video and included in product descriptions.

Product descriptions, transcripts, and all text must be available in English, Spanish, and Portuguese.

Step 1: Upload the video to blob storage

Given a video or audio file, the file is first dropped into a Blob Storage. T

Step 2: Index the video by using the Video Indexer API.

When a video is indexed, Video Indexer produces the JSON content that contains details of the specified video insights. The insights include: transcripts, OCRs, faces, topics, blocks, etc.

Step 3: Extract the transcript from the Video Indexer API.

Step 4: Translate the transcript by using the Translator API.

Parameter type: stream

Visual features list: VisualFeatureTypes.Description

Result to use: results.Description.Captions[0]

To generate accessible alt text , you should use the caption produced by Azure Computer Vision’s Description feature (it produces a human-readable sentence with a confidence score). Therefore:

The image input should be a stream , because you’re uploading images (not just passing URLs) during product creation and AnalyzeImage…Async supports image streams .

Request the Description feature in VisualFeatureTypes so the service returns results.Description.Captions.

Use results.Description.Captions[0] and return the caption text if its confidence is high enough (e.g., > 0.5) to meet the accessibility requirement that all images must have relevant alt text .

Other features (Tags, Objects, Brands) are useful for enrichment but do not directly return natural-language captions suitable for alt text.

Microsoft Azure AI Solution References

Computer Vision (Image Analysis) – Description/Captions and features: Microsoft Docs, Image Analysis – VisualFeatureTypes.Description returns description.captions.

https://learn.microsoft.com/azure/ai-services/computer-vision/concept-image-analysis

SDK usage (analyze image from a stream): Microsoft Docs, Analyze an image by using the Computer Vision client library .

https://learn.microsoft.com/azure/ai-services/computer-vision/how-to/call-analyze-image?tabs=version-3-2#analyze-an-image-from-a-stream

Comprehensive Detailed Explanation

The requirement is to build an Adaptive Card for the chatbot that supports multilingual product display and meets accessibility requirements.

Product Name Display

The product JSON stores localized names like " name " : { " en " : " ... " , " es " : " ... " , " pt " : " ... " } .

To dynamically render based on the current language, you must use name[language].

This ensures the product name shown in the card adapts automatically to the user’s preferred language.

Stock Level Warning

Business rules require warnings when stock is low or out of stock.

This means you display the stock warning only when stockLevel is not ' OK ' .

Therefore, the correct condition is:

" $when " : " ${stockLevel != ' OK ' } "

Image Alt Text for Accessibility

Accessibility requirements specify that all images must have alt text in English, Spanish, and Portuguese.

In the product JSON, " alttext " is also localized:

" altText " : { " en " : " Bicycle " , " es " : " Bicicleta " , " pt " : " Bicicleta " }

To render the correct localized alt text dynamically, use image.altText[language].

Correct Selections:

First blank: name[language]

Second blank: " $when " : " ${stockLevel != ' OK ' } "

Third blank: image.altText[language]

Microsoft References

Adaptive Cards Templating

Adaptive Card $when property

Multilingual data in Azure Cognitive Search and Adaptive Cards

Box 1: QnA Maker Editor

Scenario: Provide all employees with the ability to edit Q & As.

The QnA Maker Editor (read/write) has the following permissions:

Create KB API

Update KB API

Replace KB API

Replace Alterations

" Train API " [in

new service model v5]

Box 2: Contributor

Scenario: Only senior managers must be able to publish updates.

Contributor permission: All except ability to add new members to roles

Box 1: api-nam.cognitive.microsofttranslator.com

https://docs.microsoft.com/en-us/azure/cognitive-services/translator/reference/v3-0-reference

Box 2: /translate

C: Analyze Receipt - Get Analyze Receipt Result.

Query the status and retrieve the result of an Analyze Receipt operation.

Request URL: https://{endpoint}/formrecognizer/v2.0-preview/prebuilt/receipt/analyzeResults/{resultId}

E: POST {Endpoint}/vision/v3.2/read/analyze

Use this interface to get the result of a Read operation, employing the state-of-the-art Optical Character Recognition (OCR) algorithms optimized for text-heavy documents.

Scenario: Contoso plans to develop a document processing workflow to extract information automatically from PDFs and images of financial documents

The document processing solution must be able to process standardized financial documents that have the following characteristics:

- Contain fewer than 20 pages.

- Be formatted as PDF or JPEG files.

- Have a distinct standard for each office.

*The document processing solution must be able to extract tables and text from the financial documents.

The document processing solution must be able to extract information from receipt images.

Scenario: When the response confidence score is low, ensure that the chatbot can provide other response options to the customers.

When no good match is found by the ranker, the confidence score of 0.0 or " None " is returned and the default response is " No good match found in the KB " . You can override this default response in the bot or application code calling the endpoint. Alternately, you can also set the override response in Azure and this changes the default for all knowledge bases deployed in a particular QnA Maker service.

Choosing Ranker type: By default, QnA Maker searches through questions and answers. If you want to search through questions only, to generate an answer, use the RankerType=QuestionOnly in the POST body of the GenerateAnswer request.

The requirements state:

All AI solution responses must have a confidence score ≥ 70%.

If the response confidence score is < 70%, the response must be improved by human input.

Members of the Consultant-Bookkeeper group must be able to process financial documents, which includes performing manual review when the AI confidence is below threshold.

Looking at the provided JSON:

" fields " : {

" ReceiptType " : {

" type " : " string " ,

" valueString " : " Itemized " ,

" confidence " : 0.672

},

" MerchantName " : {

" type " : " string " ,

" valueString " : " Tailwind " ,

" confidence " : 0.913

}

}

ReceiptType.confidence = 0.672 → this is below 0.7.

MerchantName.confidence = 0.913 → this is above 0.7.

Therefore, the correct condition for triggering manual review is:

documentResults.fields. < field > .confidence < 0.7

Now, evaluating the options:

A. documentResults.docType == " prebuilt:receipt "

Always true for receipts, does not check confidence. Not correct.

B. documentResults.fields.*.confidence < 0.7

This is the correct general expression: trigger manual review whenever any field confidence is below 0.7.

C. documentResults.fields.ReceiptType.confidence > 0.7

This would bypass manual review when ReceiptType has high confidence. The requirement is to trigger review when confidence < 0.7, so this is the opposite.

D. documentResults.fields.MerchantName.confidence < 0.7

This only checks one field (MerchantName). In the JSON, MerchantName has confidence 0.913 ( > 0.7), so this condition would not trigger, but ReceiptType clearly needs review. Too narrow.

Correct Answer: B. documentResults.fields.*.confidence < 0.7

Microsoft References

Azure AI Document Intelligence – Confidence scores

Human-in-the-loop for Document Intelligence

Box 1: Cognitive Service User

Ensure that the members of a group named Management-Accountants can approve the FAQs.

Approve=publish.

Cognitive Service User (read/write/publish): API permissions: All access to Cognitive Services resource except for ability to:

1. Add new members to roles.

2. Create new resources.

Box 2: Cognitive Services QnA Maker Editor

Ensure that the members of a group named Consultant-Accountants can create and amend the FAQs.

QnA Maker Editor: API permissions:

1. Create KB API

2. Update KB API

3. Replace KB API

4. Replace Alterations

5. " Train API " [in new service model v5]

Box 3: Cognitive Services QnA Maker Read

Ensure that the members of a group named the Agent-CustomerServices can browse the FAQs.

QnA Maker Read: API Permissions:

1. Download KB API

2. List KBs for user API

3. Get Knowledge base details

4. Download Alterations

Generate Answer

Step 1: Create a new project and load sample documents

Create a new project. Projects store your configurations and settings.

Step 2: Label the sample documents

When you create or open a project, the main tag editor window opens.

Step 3: Train a custom model.

Finally, train a custom model.

Can search content in different formats, including video

Audio and video insights (multi-channels). When indexing by one channel, partial result for those models will be available.

Keywords extraction: Extracts keywords from speech and visual text.

Named entities extraction: Extracts brands, locations, and people from speech and visual text via natural language processing (NLP).

Topic inference: Makes inference of main topics from transcripts. The 2nd-level IPTC taxonomy is included.

Artifacts: Extracts rich set of " next level of details " artifacts for each of the models.

Sentiment analysis: Identifies positive, negative, and neutral sentiments from speech and visual text.

Scenario: All planned projects must support English, French, and Portuguese.

If a bot uses multiple LUIS models and QnA Maker knowledge bases (knowledge bases), you can use the Dispatch tool to determine which LUIS model or QnA Maker knowledge base best matches the user input. The dispatch tool does this by creating a single LUIS app to route user input to the correct model.

Within a search service, synonym maps are a global resource that associate equivalent terms, expanding the scope of a query without the user having to actually provide the term. For example, assuming " dog " , " canine " , and " puppy " are mapped synonyms, a query on " canine " will match on a document containing " dog " .

Create synonyms: A synonym map is an asset that can be created once and used by many indexes.

The wiki contains text in English, French and Portuguese.

Scenario: All planned projects must support English, French, and Portuguese.

The Document Extraction skill extracts content from a file within the enrichment pipeline. This allows you to take advantage of the document extraction step that normally happens before the skillset execution with files that may be generated by other skills.

Note: The Translator Text API will be used to determine the from language. The Language detection skill is not required.

Box 1: " categories " : [ " Locations " , " Persons " , " Organizations " ],

Locations, Persons, Organizations are in the outputs.

Scenario: Contoso plans to develop a searchable knowledgebase of all the intellectual property

Note: The categories parameter is an array of categories that should be extracted. Possible category types: " Person " , " Location " , " Organization " , " Quantity " , " Datetime " , " URL " , " Email " . If no category is provided, all types are returned.

Box 2: { " name " : " entities " }

The include wikis, so should include entities in the outputs.

Note: entities is an array of complex types that contains rich information about the entities extracted from text, with the following fields

name (the actual entity name. This represents a " normalized " form)

wikipediaId

wikipediaLanguage

wikipediaUrl (a link to Wikipedia page for the entity)

etc.