Azure OpenAI

WatchThatSite integrates with Azure OpenAI Service, offering enterprise-grade security and compliance for OpenAI models.

Prerequisites

• Azure account with OpenAI Service access
• Deployed Azure OpenAI models in your region

Get Your Credentials

1. Access Azure Portal

1. Go to Azure Portal
2. Navigate to Azure OpenAI resource
3. Go to Keys and Endpoint in the left sidebar

2. Copy Required Values

You'll need three pieces of information:

API Key

• Under "Key1" or "Key2", click the copy icon
• This is your authentication credential

Endpoint URL

• Copy the value shown as "Endpoint"
• Format: https://[resource-name].openai.azure.com/

API Version

• Azure uses versioned API endpoints
• Format: 2024-02-15-preview (example, check current docs)
• Find available versions in Azure OpenAI API docs

3. Verify Deployed Models

Go to Model deployments in your Azure OpenAI resource:

1. Check which models are deployed
2. Note the deployment names (they may differ from model names)
3. These deployment names are used to access the models

Add to WatchThatSite

1. Go to Settings > AI Providers
2. Click Add Provider Key
3. Select Azure OpenAI from the provider dropdown
4. Enter:
   • API Key: Your Azure OpenAI API key
   • Endpoint: Your Azure OpenAI endpoint URL
   • API Version: The API version (e.g., 2024-02-15-preview)
5. Optionally select specific models
6. Click Save

Recommended Models

Latest Azure OpenAI models with tool calling:

• GPT-4o — Latest multimodal, vision + tool calling
• GPT-4 Turbo — Fast reasoning, 128K context
• GPT-4 — Powerful general reasoning

Model availability depends on region and your deployments.

All Available Models

Your account has access to all deployed models supporting tool calling:

1. Go to Settings > AI Providers
2. Click Add Provider Key or Edit existing
3. Models dropdown shows all deployed models
4. Select specific models to restrict, or leave empty for all

Advantages

• Enterprise security: Network isolation, private endpoints
• Compliance: HIPAA, PCI, SOC 2 compliance options
• Availability: SLA guarantees and failover support
• Virtual networks: VNet support for private connectivity
• Audit logging: Detailed activity logging

Troubleshooting

"Unauthorized" Error

• Verify API key is correct
• Check key hasn't expired
• Ensure you're using the correct endpoint for your region

"Invalid Endpoint" Error

• Verify endpoint URL format: https://[resource-name].openai.azure.com/
• Ensure no trailing slashes
• Check resource name is correct

"Invalid API Version" Error

• Verify API version format is correct (e.g., 2024-02-15-preview)
• Check Azure OpenAI API docs for valid versions
• Some versions may not be available in your region

"Model Not Found"

• Verify the model is deployed in your Azure OpenAI resource
• Check deployment name matches model request
• Ensure you have quota for the model
• Some models have regional availability restrictions

Access Denied with VNet

If using VNet integration:

• Add WatchThatSite IP to allowlist
• Check network rules permit outbound traffic
• Contact your network admin for firewall rules

API Version Selection

Azure uses versioned APIs. Use the latest stable version available in your region. Check Azure docs for:

• Latest stable versions
• Deprecated versions to avoid
• Feature availability per version