Challenge 01: Select the Right Azure AI Service

Estimated Time

45 min | Cost: ~$0.50 | Domain: Plan & Manage AI Solutions (20-25%)

Exam skills covered

Select the appropriate Azure AI service for a generative AI solution
Select the appropriate Azure AI service for a computer vision solution
Select the appropriate Azure AI service for a natural language processing solution
Select the appropriate Azure AI service for a speech solution
Select the appropriate Azure AI service for a document intelligence solution
Select the appropriate Azure AI service for a knowledge mining solution

Overview

Azure AI services provide a broad portfolio of cognitive capabilities through pre-built APIs and customizable models. Choosing the right service is critical—using Azure OpenAI for simple text extraction when Document Intelligence exists, or using Computer Vision for tasks better suited to GPT-4o multimodal, leads to unnecessary cost and complexity.

This challenge walks you through the Azure AI service taxonomy, helps you build a mental decision tree, and verifies your ability to programmatically discover and validate available services in a subscription. You'll compare multi-service resources (which provide a single endpoint for multiple capabilities) against single-service resources (which offer service-specific features and isolation).

Understanding the tradeoffs between service types—pricing tiers, regional availability, feature sets, and SLA differences—is essential for the AI-102 exam and real-world architecture decisions.

Architecture

You'll create both a multi-service Azure AI resource and individual single-service resources, then programmatically enumerate their capabilities and compare their endpoints.

Challenge 01 topology

Prerequisites

Azure subscription with Azure AI services access
Azure CLI 2.50+ installed
Python 3.9+ with pip or .NET 8 SDK
azure-identity and azure-mgmt-cognitiveservices Python packages (or equivalent NuGet)

Implementation

Task 1: Create a Multi-Service Azure AI Resource

Python SDK
C# SDK
REST API

from azure.identity import DefaultAzureCredential
from azure.mgmt.cognitiveservices import CognitiveServicesManagementClient
from azure.mgmt.cognitiveservices.models import Account, Sku, AccountProperties

credential = DefaultAzureCredential()
subscription_id = "YOUR_SUBSCRIPTION_ID"
client = CognitiveServicesManagementClient(credential, subscription_id)

# Create a multi-service resource
account = client.accounts.begin_create(
    resource_group_name="rg-ai102-challenge01",
    account_name="ai-multiservice-01",
    account=Account(
        sku=Sku(name="S0"),
        kind="AIServices",
        location="eastus",
        properties=AccountProperties()
    )
).result()

print(f"Created: {account.name}")
print(f"Endpoint: {account.properties.endpoint}")
print(f"Kind: {account.kind}")

using Azure.Identity;
using Azure.ResourceManager;
using Azure.ResourceManager.CognitiveServices;
using Azure.ResourceManager.CognitiveServices.Models;

var credential = new DefaultAzureCredential();
var client = new ArmClient(credential);

var subscription = await client.GetDefaultSubscriptionAsync();
var resourceGroup = await subscription
    .GetResourceGroupAsync("rg-ai102-challenge01");

var collection = resourceGroup.Value.GetCognitiveServicesAccounts();

var data = new CognitiveServicesAccountData(Azure.Core.AzureLocation.EastUS)
{
    Kind = "CognitiveServices",
    Sku = new CognitiveServicesSku("S0"),
    Properties = new CognitiveServicesAccountProperties()
};

var result = await collection.CreateOrUpdateAsync(
    Azure.WaitUntil.Completed, "ai-multiservice-01", data);

Console.WriteLine($"Created: {result.Value.Data.Name}");
Console.WriteLine($"Endpoint: {result.Value.Data.Properties.Endpoint}");

# Create multi-service resource
az cognitiveservices account create \
  --name ai-multiservice-01 \
  --resource-group rg-ai102-challenge01 \
  --kind AIServices \
  --sku S0 \
  --location eastus \
  --yes

# Get endpoint and keys
az cognitiveservices account show \
  --name ai-multiservice-01 \
  --resource-group rg-ai102-challenge01 \
  --query "properties.endpoint" -o tsv

az cognitiveservices account keys list \
  --name ai-multiservice-01 \
  --resource-group rg-ai102-challenge01

Task 2: List Available AI Service Kinds

Python SDK
C# SDK
REST API

# List all available cognitive service kinds in the subscription
kinds = client.resource_skus.list()
service_kinds = set()
for sku in kinds:
    service_kinds.add(sku.kind)

print("Available Azure AI service kinds:")
for kind in sorted(service_kinds):
    print(f"  - {kind}")

# Key kinds for AI-102:
# CognitiveServices (multi-service), OpenAI, ComputerVision,
# TextAnalytics, SpeechServices, FormRecognizer, ContentSafety

using Azure.Identity;
using Azure.ResourceManager;
using Azure.ResourceManager.CognitiveServices;

var credential = new DefaultAzureCredential();
var client = new ArmClient(credential);
var subscription = await client.GetDefaultSubscriptionAsync();

// List available SKUs to see service kinds
var skus = subscription.GetCognitiveServicesResourceSkusAsync();
var serviceKinds = new HashSet<string>();

await foreach (var sku in skus)
{
    serviceKinds.Add(sku.Kind);
}

Console.WriteLine("Available Azure AI service kinds:");
foreach (var kind in serviceKinds.OrderBy(k => k))
{
    Console.WriteLine($"  - {kind}");
}

# List all cognitive services accounts in a resource group
az cognitiveservices account list \
  --resource-group rg-ai102-challenge01 \
  -o table

# List available kinds in a region
az cognitiveservices account list-skus \
  --kind AIServices \
  --location eastus \
  -o table

# Check specific service availability
az cognitiveservices account list-skus \
  --kind OpenAI \
  --location eastus \
  -o table

Task 3: Create Single-Service Resources and Compare

Python SDK
C# SDK
REST API

# Create individual service resources for comparison
services = [
    {"name": "ai-vision-01", "kind": "ComputerVision", "sku": "S1"},
    {"name": "ai-language-01", "kind": "TextAnalytics", "sku": "S"},
    {"name": "ai-speech-01", "kind": "SpeechServices", "sku": "S0"},
]

for svc in services:
    result = client.accounts.begin_create(
        resource_group_name="rg-ai102-challenge01",
        account_name=svc["name"],
        account=Account(
            sku=Sku(name=svc["sku"]),
            kind=svc["kind"],
            location="eastus",
            properties=AccountProperties()
        )
    ).result()
    print(f"Created {svc['kind']}: {result.properties.endpoint}")

# Compare: multi-service has ONE endpoint for all
# Single-service has dedicated endpoints with service-specific features
multi = client.accounts.get("rg-ai102-challenge01", "ai-multiservice-01")
print(f"\nMulti-service endpoint: {multi.properties.endpoint}")
print("Supports: Vision, Language, Speech, Decision (single key)")

var servicesToCreate = new[]
{
    new { Name = "ai-vision-01", Kind = "ComputerVision", Sku = "S1" },
    new { Name = "ai-language-01", Kind = "TextAnalytics", Sku = "S" },
    new { Name = "ai-speech-01", Kind = "SpeechServices", Sku = "S0" }
};

foreach (var svc in servicesToCreate)
{
    var svcData = new CognitiveServicesAccountData(Azure.Core.AzureLocation.EastUS)
    {
        Kind = svc.Kind,
        Sku = new CognitiveServicesSku(svc.Sku),
        Properties = new CognitiveServicesAccountProperties()
    };

    var result = await collection.CreateOrUpdateAsync(
        Azure.WaitUntil.Completed, svc.Name, svcData);
    Console.WriteLine($"Created {svc.Kind}: {result.Value.Data.Properties.Endpoint}");
}

# Create individual services
az cognitiveservices account create \
  --name ai-vision-01 \
  --resource-group rg-ai102-challenge01 \
  --kind ComputerVision --sku S1 --location eastus --yes

az cognitiveservices account create \
  --name ai-language-01 \
  --resource-group rg-ai102-challenge01 \
  --kind TextAnalytics --sku S --location eastus --yes

az cognitiveservices account create \
  --name ai-speech-01 \
  --resource-group rg-ai102-challenge01 \
  --kind SpeechServices --sku S0 --location eastus --yes

# Compare endpoints
echo "Multi-service endpoint:"
az cognitiveservices account show --name ai-multiservice-01 \
  --resource-group rg-ai102-challenge01 --query "properties.endpoint" -o tsv

echo "Vision endpoint:"
az cognitiveservices account show --name ai-vision-01 \
  --resource-group rg-ai102-challenge01 --query "properties.endpoint" -o tsv

Expected Output

Created: ai-multiservice-01
Endpoint: https://eastus.api.cognitive.microsoft.com/
Kind: CognitiveServices

Available Azure AI service kinds:
  - CognitiveServices
  - ComputerVision
  - ContentSafety
  - FormRecognizer
  - OpenAI
  - SpeechServices
  - TextAnalytics
  ...

Created ComputerVision: https://eastus.api.cognitive.microsoft.com/
Created TextAnalytics: https://eastus.api.cognitive.microsoft.com/
Created SpeechServices: https://eastus.cognitiveservices.azure.com/

Multi-service endpoint: https://eastus.api.cognitive.microsoft.com/
Supports: Vision, Language, Speech, Decision (single key)

Break & fix

Scenario	Symptom	Root Cause	Fix
Wrong kind specified	`InvalidParameterValue` error	Using deprecated kind name (e.g., "Face" vs "CognitiveServices")	Check `az cognitiveservices account list-skus` for valid kinds
Region not available	`LocationNotAvailable` error	Service not available in chosen region	Use `az account list-locations` and check service availability matrix
SKU mismatch	`SkuNotAvailable`	Requested SKU not offered for that kind	Match SKU to service kind (e.g., TextAnalytics uses "S" not "S0")
Quota exceeded	`QuotaExceeded`	Too many resources of same kind in subscription	Delete unused resources or request quota increase

Knowledge Check

1. You need to use a single endpoint and key to access Computer Vision, Language, and Speech capabilities. Which resource kind should you create?

2. Which Azure AI service should you use to extract structured data from invoices and receipts?

3. What is a key limitation of multi-service Azure AI resources compared to single-service resources?

4. You need to implement real-time speech translation for a conference application. Which service should you select?

5. Which scenario requires Azure OpenAI Service rather than Azure AI Language?

Cleanup

az group delete --name rg-ai102-challenge01 --yes --no-wait

Exam skills covered​

Overview​

Architecture​

Prerequisites​

Implementation​

Task 1: Create a Multi-Service Azure AI Resource​

Task 2: List Available AI Service Kinds​

Task 3: Create Single-Service Resources and Compare​

Expected Output​

Break & fix​

Knowledge Check​

Cleanup​

Learn More​