AI-102 Exam Dumps - Designing and Implementing a Microsoft Azure AI Solution

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.

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.

Review of Answer Options:

Correct Answer: Distributes processing across compute nodes

Azure Synapse Analytics architecture

Distributed query processing in Azure Synapse

Microsoft References

Upload a glossary file to the container for French files

Define a document translation specification that has a French target

Perform an asynchronous translation by using the document translation specification

You need to translate Word and PowerPoint files (German → French) while preserving formatting and supporting a custom glossary. The correct service is Azure Document Translation, part of the Translator service in Azure Cognitive Services.

Step-by-step reasoning:

Upload a glossary file to the container for French files.

Glossaries (also called custom terminology files) guide translation to ensure consistent word usage.

They are stored in the target container (in this case, the French container).

Define a document translation specification that has a French target.

A translation request is defined by creating a document translation specification JSON.

This specification indicates:

The source container (German files)

The target container (French files)

The glossary file (for custom terminology)

Output target language (French).

Perform an asynchronous translation by using the document translation specification.

Document translation jobs are submitted asynchronously.

This step executes the translation while preserving formatting and applying the glossary rules.

Why not the other options?

Generate a list of files to be translated → not required, the service automatically processes all files in the source container.

Perform asynchronous translation using a list of files → that method does not support document-level translation with glossary and formatting preservation, it’s used for text-based translations.

Correct Answer Order:

Upload a glossary file to the container for French files

Define a document translation specification that has a French target

Perform an asynchronous translation by using the document translation specification

Document Translation – Translate documents in containers

Use custom glossaries with Document Translation

Submit document translation requests

Microsoft References

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

You have files with the following formats:

File1: WMV (34 mins, 400 MB)

File2: AVI (90 mins, 1,200 MB)

File3: MOV (300 mins, 980 MB)

File4: MP4 (80 mins, 1,800 MB)

You must determine which files can be uploaded to Azure AI Video Indexer.

Supported formats: MP4, MOV, WMV, AVI, M2TS.

Maximum file length: 4 hours (240 minutes).

Maximum file size: 2 GB (2048 MB) for website uploads.

Key Rules (Video Indexer upload limitations):

File1 (WMV, 34 min, 400 MB): Supported format, within time and size → ✅ Allowed.

File2 (AVI, 90 min, 1,200 MB): Supported format, within time and size → ✅ Allowed.

File3 (MOV, 300 min, 980 MB): Supported format, but exceeds 240 min limit → ❌ Not Allowed.

File4 (MP4, 80 min, 1,800 MB): Supported format, within size (under 2 GB), within time → ✅ Allowed.

Check each file:

Correct Answer:

A. File1, File2, and File4 only

Microsoft Reference:

Video Indexer upload limitations

 Service endpoint → Computer Vision analyze images

 Property → imageType

The task:

"You have a library with thousands of images. You need to tag the images as photographs, drawings, or clipart."

Which service endpoint?

Computer Vision analyze images → Provides a broad analysis including categories, tags, descriptions, and image type (photo, drawing, clipart). ✅

Computer Vision object detection → Identifies objects (people, cars, animals), not image type.

Custom Vision image classification / object detection → Requires custom training, not needed since the built-in Computer Vision service already supports this.

Correct: Computer Vision analyze images

Which property?

categories → Provides predefined categories (e.g., outdoor, indoor, people), not specifically drawing/clipart/photo.

description → Generates captions in natural language, not classification into photo/drawing/clipart.

imageType → Identifies if an image is a photograph, line drawing, or clipart. ✅

metadata → Provides size, format, dimensions.

objects → Lists detected objects, not image type.

Correct: imageType

Correct Answer:

Service endpoint → Computer Vision analyze images

Property → imageType

Computer Vision Analyze API

ImageType property in Computer Vision

Microsoft References

Database transactions follow the ACID properties:

Atomicity → Ensures that each transaction is treated as a single “all-or-nothing” unit of work. Either the entire transaction is completed, or none of it is applied. This guarantees that transactions are executed only once and rolled back completely if they fail.

Consistency → Ensures that a transaction brings the database from one valid state to another, preserving integrity constraints.

Isolation → Ensures that concurrent transactions do not interfere with each other.

Durability → Ensures that once a transaction is committed, it remains permanent even in case of system failure.

Thus, the correct property here is Atomicity.

Microsoft Reference: ACID properties in databases

Scenario RecapYou are building a streaming Speech-to-Text solution using the Speech SDK.

Input audio format: MP3 encoding

You must configure the audio stream format and the recognizer properly.

Audio Format

For MP3 streams, the correct method is:

AudioStreamFormat.GetCompressedFormat(AudioStreamContainerFormat.MP3)

This creates an audio format object for compressed MP3 input.

Recognizer

Since the task is speech-to-text transcription, you need the SpeechRecognizer class.

Other options like KeywordRecognizer, SpeakerRecognizer, or SpeechSynthesizer are not appropriate (they handle keyword spotting, speaker identification, or text-to-speech).

Code Analysis

Completed Code Snippetvar audioFormat = AudioStreamFormat.GetCompressedFormat(AudioStreamContainerFormat.MP3);

var speechConfig = SpeechConfig.FromSubscription("18c51a87-3a69-47a8-aedc-a54745f708a1", "westus");

var audioConfig = AudioConfig.FromStreamInput(pushStream, audioFormat);

using (var recognizer = new SpeechRecognizer(speechConfig, audioConfig))

{

var result = await recognizer.RecognizeOnceAsync();

var text = result.Text;

}

AudioFormat: AudioStreamFormat.GetCompressedFormat

Recognizer: SpeechRecognizer

Final Answer (Answer Area Selections)

Speech SDK – AudioStreamFormat

SpeechRecognizer class

Microsoft References

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)

Custom Vision project types: Classification vs Object detection

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

Exporting compact models for offline use

Microsoft References