AI-102 Practice Questions

Comprehensive Detailed Explanation

You are deploying Anomaly Detector API as a container. The key requirements are:

Prevent billing and API keys from being exposed in command-line history.

Control access to container images using Azure RBA

C.  

Step-by-step reasoning:

Pull the Anomaly Detector container image

Microsoft Cognitive Services (including Anomaly Detector) provides container images via Microsoft Container Registry.

You must first pull the container image locally.

Push the image to an Azure container registry (ACR)

To enforce Azure RBAC, you need to use Azure Container Registry instead of Docker Hub.

Storing in ACR ensures authentication and role-based access control to the image.

Distribute a docker run script

Instead of providing credentials directly on the command line (which risks storing API keys in shell history), best practice is to distribute a docker run script (or use environment variables / mounted files) to run the container.

This prevents exposure of sensitive information.

Run the container on local devices or on-premises datacenters

The container now runs securely, with API key and billing info injected at runtime via environment variables or mounted secrets.

Incorrect Options

Create a custom Dockerfile → Not needed, because Microsoft provides the official Anomaly Detector container image.

Build the image → Not required; again, the container is prebuilt.

Push the image to Docker Hub → Does not allow RBAC, so this would not meet the security requirement.

Correct Sequence:

Pull the Anomaly Detector container image.

Push the image to an Azure container registry.

Distribute a docker run script.

Run the container in the local or on-prem environment.

Microsoft References

Use Cognitive Services containers

Azure Container Registry – Authentication and RBAC

Best practices for running Cognitive Services containers

In relational database systems such as Microsoft SQL Server or Azure SQL Database, the following definitions apply:

Stored Procedure:

A stored procedure is a precompiled block of Transact-SQL (T-SQL) code stored in the database. It can contain control-of-flow statements, queries, and business logic. When executed, it runs as a single unit within the database engine. This matches the definition of " a block of code that runs in a database. "

Table:

A table is a structured set of rows and columns used to store data, not a block of executable code.

View:

A view is a virtual table based on the result of a SQL query. It is used for abstraction and simplification of queries, not for executing code blocks.

Index:

An index is a database structure that improves the speed of data retrieval. It is not executable code.

Thus, the only option that correctly describes " a block of code that runs in a database " is a stored procedure.

Correct Answer: A stored procedure

Microsoft References

Stored Procedures (Database Engine)

Azure SQL Database - Stored procedures overview

SQL Server Views

SQL Server Tables

SQL Server Indexes

Text Extraction: b) Computer Vision

Computer Vision is a service that can analyze images and extract text from them using optical character recognition (OCR) or read API1. OCR can detect and extract printed or handwritten text from images in various languages and formats, such as PDF, TIFF, or JPEG2. Read API can perform asynchronous batch processing of text from multiple images, and can also handle text in tables, mixed languages, and rotated text3.

Profane Language Detection: c) Content Moderator

Content Moderator is a service that can detect and filter out potentially offensive or unwanted content in text, images, and videos. For text, Content Moderator can identify profane or abusive language, personal data, and custom terms that you define. You can also use Content Moderator to review and moderate the content manually or automatically using workflows.

For deployment to a mobile app (Android), Custom Vision requires a compact domain so the model can be exported and run on-device (e.g., as TensorFlow Lite or ONNX). The model was trained using the Retail domain, which is not an exportable compact domain for object detection. Therefore:

Change the model domain to an exportable Object Detection (compact) domain (e.g., General [compact]).

Retrain the model so weights are learned under the new domain—domain changes require retraining.

Export the model for Android (typically TensorFlow Lite or ONNX) and integrate it into the app.

This sequence prepares the model for on-device mobile deployment with minimal extra development.

References (Microsoft Azure AI Solution)

Custom Vision: Only compact domains support model export for edge/mobile.

Changing a domain requires retraining the iteration.

Exporting object detection models supports TensorFlow/TensorFlow Lite/ONNX formats for Android integration.

You are developing a speech translation service for recording lectures given in English (United Kingdom). The goal is to provide transcripts in multiple target languages: English, French, Spanish, and German.

Key requirements from the code snippet:

Source Language (Recognition Language):

config.SpeechRecognitionLanguage = " en-GB " ;

The input is English (UK).

Target Languages:

You need to define a list of supported target languages. Since English, French, Spanish, and German are supported, the correct syntax is:

var lang = new List < string > () { " en " , " fr " , " de " , " es " };

This uses ISO language codes expected by the Translator API ( " fr " = French, " de " = German, " es " = Spanish).

Recognizer Type:

Since this is speech translation, you must use:

using var recognizer = new TranslationRecognizer(config, audioConfig);

The TranslationRecognizer class is specifically for translating speech input into multiple languages.

Completed Code

static async Task TranslateSpeechAsync ()

{

var config = SpeechTranslationConfig.FromSubscription( " 69cad5cc-0ab3-4704-bdff-abf4aa07d85 " , " uksouth " );

var lang = new List < string > () { " en " , " fr " , " de " , " es " };

config.SpeechRecognitionLanguage = " en-GB " ;

lang.ForEach(config.AddTargetLanguage);

using var audioConfig = AudioConfig.FromDefaultMicrophoneInput();

using var recognizer = new TranslationRecognizer(config, audioConfig);

var result = await recognizer.RecognizeOnceAsync();

if (result.Reason == ResultReason.TranslatedSpeech)

{

foreach ( var element in result.Translations)

{

AppendToTranscriptFile(element.Value, element.Key);

}

}

}

Final Answer (Answer Area Selections)

First Blank: { " en " , " fr " , " de " , " es " }

Second Blank: TranslationRecognizer

Microsoft References

Speech Translation with Speech SDK

TranslationRecognizer class

Speech service supported languages

a tabular form of rows and columns

Rationale

Relational data is, by definition, data that adheres to the relational model .

In the relational model, data is organized into tables , which consist of rows (records or tuples) and columns (fields or attributes). This is known as the tabular form .

The other options are incorrect:

Unstructured data is the opposite of relational data.

A hierarchical folder structure is used for file organization, not the logical structure of relational data.

Comma-separated value (CSV) files can contain relational data, but the core definition of relational data is the tabular structure , not the file format itself.

Scenario Recap

You are building a remote training monitoring solution.

Requirement: Use video and audio feeds to detect if a learner is present, paying attention, and talking.

Services available: Face, Speech, Text Analytics.

Analysis

From a learner’s video feed, verify whether the learner is present.

The Face API can detect and identify faces in a video feed.

It can tell if a person is present and recognized, fulfilling the requirement.

From a learner’s facial expression in the video feed, verify whether the learner is paying attention.

Again, the Face API provides facial expression and emotion recognition (happiness, anger, neutral, etc.).

This can be mapped to “paying attention vs. distracted.”

From a learner’s audio feed, detect whether the learner is talking.

The Speech service detects spoken input and can determine if speech is present.

Text Analytics works on text (not raw audio) and is therefore not appropriate here.

Final Answer (Answer Area Selections)

From a learner’s video feed, verify whether the learner is present: Face

From a learner’s facial expression in the video feed, verify whether the learner is paying attention: Face

From a learner’s audio feed, detect whether the learner is talking: Speech

Microsoft References

Face API – Face detection & identification

Face API – Emotion recognition

Azure Speech service

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.

The chatbot needs to recognize two different kinds of information:

Code names for internal product development

These are not standard entities like " locations " or " organizations " that prebuilt NER can detect.

You need to train a custom model to recognize internal terms or codenames (for example, " Project Falcon " ).

The correct Azure Cognitive Service for Language feature is Custom Named Entity Recognition (Custom NER), which allows defining and training entity categories specific to your business.

Messages that include credit card numbers

Credit card numbers are sensitive data falling under Personally Identifiable Information (PII).

Azure Cognitive Service for Language provides a PII detection feature that automatically identifies and masks sensitive information such as credit card numbers, SSNs, and phone numbers.

This minimizes development effort since it is prebuilt and ready to use.

Correct Answer Mapping:

Identify code names for internal product development → Custom named entity recognition (NER)

Identify messages that include credit card numbers → Personally Identifiable Information (PII) detection

Microsoft References

Custom Named Entity Recognition in Azure AI Language

PII detection in Azure AI Language