docs: APRL recommendations for Cog Services (#575)

Co-authored-by: Mandar Gujrathi <[email protected]>
Co-authored-by: Zach Trocinski <[email protected]>

azure-resources/CognitiveServices/accounts/kql/081fc8a4-b2d9-405b-b351-334e621016f5.kql

@@ -0,0 +1 @@
+// cannot-be-validated-with-arg

azure-resources/CognitiveServices/accounts/recommendations.yaml

@@ -1,3 +1,19 @@
+- description: Ensure AOAI models are deployed using Global deployment
+  aprlGuid: 081fc8a4-b2d9-405b-b351-334e621016f5
+  recommendationTypeId: null
+  recommendationControl: HighAvailability
+  recommendationImpact: High
+  recommendationResourceType: Microsoft.CognitiveServices/Accounts
+  recommendationMetadataState: Active
+  longDescription: Global deployments leverage Azure's global infrastructure to route customer traffic to the best available data center for the customer’s inference requests. This ensures highest initial throughput limits and best model availability while still providing our uptime SLA and low latency.
+  potentialBenefits: Low latency, best model availability, business continuity
+  pgVerified: false
+  automationAvailable: false
+  tags: null
+  learnMoreLink:
+    - name: Learn More
+      url: "https://learn.microsoft.com/en-us/azure/ai-services/openai/how-to/deployment-types#global-provisioned"
+
 - description: Deploy a PAYG instance of the model with provisioned throughput to manage overflow effectively
   aprlGuid: 0c193899-da60-4a52-b4a0-77d75ac8c5c5
   recommendationTypeId: null