Azure OpenAI Service: Getting Started

Azure OpenAI Service

Azure OpenAI Service provides REST API access to OpenAI's powerful models (GPT-4o, GPT-4, GPT-3.5 Turbo, DALL-E, Whisper) with the enterprise security, compliance, and regional availability of Azure.

Why Azure OpenAI vs OpenAI Direct?

Feature	Azure OpenAI	OpenAI API
Data privacy	Your data stays in your Azure tenant	Sent to OpenAI servers
Compliance	SOC 2, HIPAA, ISO 27001, FedRAMP	Limited certifications
Networking	Private endpoints, VNet integration	Public API only
Content safety	Configurable content filters	Fixed safety layer
SLA	99.9% uptime SLA	Best-effort
Scale	Provisioned throughput (PTU)	Shared capacity

Prerequisites

• Azure subscription
• Azure OpenAI access (apply at aka.ms/oai/access)

Step 1: Create an Azure OpenAI Resource

# Using Azure CLI
az cognitiveservices account create \
  --name "my-openai-resource" \
  --resource-group "my-rg" \
  --kind "OpenAI" \
  --sku "S0" \
  --location "eastus2"

Or via the Azure Portal:

1. Search for "Azure OpenAI" in the marketplace
2. Click Create
3. Choose subscription, resource group, region, and name
4. Click Review + Create

Step 2: Deploy a Model

# Deploy GPT-4o
az cognitiveservices account deployment create \
  --name "my-openai-resource" \
  --resource-group "my-rg" \
  --deployment-name "gpt-4o" \
  --model-name "gpt-4o" \
  --model-version "2024-08-06" \
  --model-format "OpenAI" \
  --sku-capacity 30 \
  --sku-name "Standard"

Step 3: Use the Playground

1. Go to Azure AI Foundry or the Azure OpenAI portal
2. Select your deployment
3. Open the Chat Playground
4. Configure:
   • System message: Set the AI's persona and rules
   • Temperature: 0 (deterministic) to 1 (creative)
   • Max tokens: Limit response length
   • Top P: Nucleus sampling parameter

Example System Message

You are a helpful Azure solutions architect assistant.
You help users design cloud architectures using Azure services.
Always provide cost-effective, scalable, and secure recommendations.
Use markdown formatting and include relevant Azure service names.

Step 4: API Integration

Python SDK

from openai import AzureOpenAI

client = AzureOpenAI(
    api_key="your-api-key",
    api_version="2024-08-01-preview",
    azure_endpoint="https://my-openai-resource.openai.azure.com/"
)

response = client.chat.completions.create(
    model="gpt-4o",  # deployment name
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Explain Azure Functions in 3 sentences."}
    ],
    temperature=0.7,
    max_tokens=500
)

print(response.choices[0].message.content)

C# / .NET SDK

using Azure.AI.OpenAI;
using Azure;

var client = new AzureOpenAIClient(
    new Uri("https://my-openai-resource.openai.azure.com/"),
    new AzureKeyCredential("your-api-key"));

var chatClient = client.GetChatClient("gpt-4o");

var response = await chatClient.CompleteChatAsync(new[]
{
    new SystemChatMessage("You are a helpful assistant."),
    new UserChatMessage("Explain Azure Functions in 3 sentences.")
});

Console.WriteLine(response.Value.Content[0].Text);

Step 5: Content Filtering

Azure OpenAI includes configurable content filters:

1. Go to Content filters in the Azure OpenAI portal
2. Create a custom filter configuration
3. Set severity thresholds for: Hate, Sexual, Violence, Self-harm
4. Configure Prompt Shield to block jailbreak attempts
5. Assign the filter to your deployment

Resources

• Azure OpenAI Documentation
• Quickstart: Chat Completions
• Microsoft Learn: Azure OpenAI Training

Video: Watch the Azure OpenAI Getting Started on Microsoft Azure YouTube channel.