Merge pull request #53160 from MicrosoftDocs/main

Auto Publish – main to live - 2026-01-20 00:00 UTC

Author: learn-build-service-prod[bot]

learn-pr/paths/implement-container-app-hosting-azure/index.yml

@@ -0,0 +1,37 @@
+### YamlMime:LearningPath
+uid: learn.wwl.implement-container-app-hosting-azure
+metadata:
+  title: Implement Container Application Hosting on Azure
+  description: Learn how to deploy, configure, and troubleshoot containerized applications on Azure.
+  ms.date: 1/19/2026
+  author: jeffkoms
+  ms.author: jeffko
+  ms.topic: learning-path
+title: Implement container application hosting on Azure
+prerequisites: |
+  - Programming experience with languages such as Python, JavaScript, or C#.
+  - Basic understanding of Azure services and cloud computing concepts.
+  - Familiarity with containerization fundamentals.
+summary: |
+  This learning path guides you through core container hosting workflows on Azure for AI applications and backend services. You start by using Azure Container Registry to store and organize images, build images in the cloud with ACR Tasks, and apply tagging and versioning practices that support reliable deployments.
+
+  You then deploy custom containers to Azure App Service, configure runtime behavior such as ports, startup commands, and persistent storage, and externalize environment-specific configuration using App Service application settings.
+
+  Finally, you use Azure Container Instances to run containers quickly, set environment variables, control restart policies for task-style workloads, and mount Azure Files shares when your container needs shared storage.
+iconUrl: /training/achievements/generic-trophy.svg
+levels:
+- intermediate
+roles:
+- developer
+products:
+- azure-container-registry
+- azure-app-service
+- azure-container-instances
+subjects:
+- containers
+modules:
+- learn.wwl.store-manage-containers-azure-container-registry
+- learn.wwl.deploy-containers-azure-app-service
+- learn.wwl.create-run-container-images-azure-container-instances
+trophy:
+  uid: learn.wwl.implement-container-app-hosting-azure.trophy

learn-pr/wwl-azure/design-ai-workloads/6-governance.yml

@@ -1,6 +1,6 @@
 ### YamlMime:ModuleUnit
 uid: learn.wwl.design-ai-workloads.governance
-title: "Govern AI"
+title: "Govern AI workloads"
 metadata:
   title: "Govern AI"
   description: "Learn how to govern AI workloads responsibly by applying principles, managing risks, and ensuring security, compliance, and cost control."
@@ -10,4 +10,4 @@ metadata:
   ms.topic: unit
 durationInMinutes: 3
 content: |
-  [!include[](includes/6-governance.md)]
+  [!include[](includes/6-governance.md)]

learn-pr/wwl-azure/design-ai-workloads/7-secure-workloads.yml

@@ -1,6 +1,6 @@
 ### YamlMime:ModuleUnit
 uid: learn.wwl.design-ai-workloads.secure-workloads
-title: "Secure AI"
+title: "Secure AI workloads"
 metadata:
   title: "Secure AI"
   description: "Learn to secure AI workloads by identifying risks, protecting resources, and safeguarding data."
@@ -10,4 +10,4 @@ metadata:
   ms.topic: unit
 durationInMinutes: 5
 content: |
-  [!include[](includes/7-secure-workloads.md)]
+  [!include[](includes/7-secure-workloads.md)]

learn-pr/wwl-azure/design-ai-workloads/8-manage.yml

@@ -1,6 +1,6 @@
 ### YamlMime:ModuleUnit
 uid: learn.wwl.design-ai-workloads.manage
-title: "Manage AI"
+title: "Manage AI workloads"
 metadata:
   title: "Manage AI"
   description: "Learn how to manage AI workloads effectively, including operations, deployment, costs, and business continuity, using Azure tools and best practices."
@@ -10,4 +10,4 @@ metadata:
   ms.topic: unit
 durationInMinutes: 3
 content: |
-  [!include[](includes/8-manage.md)]
+  [!include[](includes/8-manage.md)]

learn-pr/wwl-azure/design-ai-workloads/includes/10-summary.md

@@ -22,4 +22,9 @@ Copilot can assist you in configuring Azure infrastructure solutions. Copilot ca
 
 * Summarize best practices for governing, securing, and managing AI workloads on Azure. Include key services and practices for Responsible AI, security, cost management, and operations.
 
- 
+### Learn more with Azure documentation
+
+* The [AI adoption - Cloud Adoption Framework](/azure/cloud-adoption-framework/scenarios/ai/) provides a structured process for adopting AI solutions in Azure.
+  
+* The [AI Workload Documentation - Microsoft Azure Well-Architected Framework](/azure/well-architected/ai/) focuses on the architectural challenges of designing AI workloads. 
+

learn-pr/wwl-azure/design-ai-workloads/includes/4-platform-workloads.md

@@ -29,6 +29,6 @@ Select resources that help you build AI workloads with prebuilt or custom soluti
 
 Keep your AI workloads inside a [secure virtual network](/azure/cloud-adoption-framework/scenarios/ai/platform/networking) and avoid exposing anything to the public internet. Use private endpoints so services like Foundry, Azure AI Services, Storage, Key Vault, and Container Registry all communicate privately over the Azure backbone.
 
-Control traffic tightly. Use a Bastion-connected jump box for safe admin access, restrict outbound traffic to only what your workloads need, and use network security groups to enforce least‑privilege access between resources.
+Control traffic tightly. Use Bastion for admin access, restrict access with private links, rely on solid DNS, and use NSGs to enforce least‑privilege access between resources.
 
 Finally, monitor your environment. Azure Firewall protects and inspects outbound traffic, and Azure Web Application Firewall shields any public-facing AI apps from common attacks, helping you run AI workloads securely and confidently.

learn-pr/wwl-azure/design-ai-workloads/includes/5-infrastructure-workloads.md

@@ -26,13 +26,13 @@ Use high-performance file storage for hot data actively used during AI jobs to s
 
 :::image type="content" source="../media/network-concepts.png" alt-text="Diagram of a virtual network with platform concepts.":::
 
-AI workloads can move a ton of data, so you want a connection that doesn’t slow everything down. Consider Azure ExpressRoute that gives you a dedicated, high-bandwidth path that keeps data flowing quickly and reliably.
+AI workloads can move a ton of data, so you want a connection that doesn’t slow everything down. Consider Azure ExpressRoute that gives you a dedicated, high-bandwidth path that keeps data flowing quickly and reliably. Bandwidth is crucial for cloud processing of on-premises storage.
 
 It also helps to keep latency as low as possible. Placing your VMs and resources close together—ideally in the same region or in a Proximity Placement Group—cuts down on the travel time for data and makes training jobs run more efficiently.
 
-And if you’re working with multiple GPUs or multiple VMs, fast networking between nodes becomes essential. InfiniBand-enabled Azure VMs make GPU-to-GPU communication super-fast, and tools like Azure Batch can handle the setup for you automatically.
+And if you’re working with multiple GPUs or multiple VMs, fast networking between nodes becomes essential. Consider [high performance networking](/azure/cloud-adoption-framework/scenarios/ai/infrastructure/networking) particularly for GPU-accelerated tasks.
 
-> [!NOTE]
+> [!TIP]
 >  The [Introduction to AI Landing Zones](/training/modules/intro-ai-landing-zones/) training module provides details on Azure landing zones with AI workloads.
 
 

learn-pr/wwl-azure/design-ai-workloads/includes/6-governance.md

@@ -23,6 +23,9 @@ Secure your AI environment by limiting access to only what’s necessary. Use ro
 
 Keep AI spending predictable by setting budgets and alerts in [Cost Management](/azure/cost-management-billing/costs/reporting-get-started). Optimize resources with autoscaling, virtual machine (VM) autoshutdown, and approved VM SKUs, and control model costs by choosing the right pricing approach—whether that’s pay-as-you-go, PTUs, or reservations.
 
+> [!TIP]
+> To learn more, consider the [Introduction to cost management for AI workloads](/training/modules/understand-cost-management-ai/) training module.
+
 **Data Governance & Privacy**
 
 Strong data governance starts with clear standards. Decide what data can be used for training and RAG. Protect privacy with techniques like anonymization and data residency controls, and make sure you have solid retention and deletion policies in place. Enable logging and diagnostics so you always have visibility into how data is accessed and used.
@@ -31,7 +34,7 @@ Strong data governance starts with clear standards. Decide what data can be used
 
 Use [Azure Policy](/azure/governance/policy/overview) to consistently enforce deployment and configuration standards across your AI environment. Apply structured tagging for cost tracking and ownership. Use [Microsoft Purview ](/purview/purview)to help secure and govern all your data.
 
-> [!NOTE]
+> [!TIP]
 > To learn more, consider the [Govern AI Services with Azure Policy](/training/modules/govern-ai-azure-policy/) training module. 
 
 **Operations**

learn-pr/wwl-azure/design-ai-workloads/includes/8-manage.md

@@ -1,7 +1,7 @@
 Managing AI operations is about keeping your AI work consistent and reliable. Create an [AI Center of Excellence](/azure/cloud-adoption-framework/scenarios/ai/center-of-excellence) to guide strategy, choose the right framework (MLOps or GenAIOps), and standardize SDKs and APIs. Use sandboxes for safe experimentation and simplify tuning with tools like [Copilot Tuning.](/copilot/microsoft-365/copilot-tuning-overview) 
 
 > [!TIP]
-> For more information, consider the [Introduction to AI Center of Excellence](/training/modules/intro-ai-center-excellence/) training module.
+> To learn more, consider the [Introduction to AI Center of Excellence](/training/modules/intro-ai-center-excellence/) training module.
 
 :::image type="content" source="../media/manage-operations.png" alt-text="Diagram of the main management areas.":::
 
@@ -27,3 +27,7 @@ Managing AI data means ensuring its accurate, secure, and compliant. Create and
 To keep AI services running during disruptions, you need a solid continuity plan that includes constant monitoring, multi-region deployments, and regular disaster-recovery testing. Continuous monitoring helps you catch issues early as models, data, and usage patterns evolve. 
 
 Deploying AI workloads—both generative and traditional—across multiple Azure regions provides redundancy and prevents outages. Use tools like Azure Front Door and Traffic Manager for automatic failovers. Strong version control and automated backups make it easy to restore models, pipelines, and configurations when something goes wrong. 
+
+> [!IMPORTANT]
+> Be sure to document your recovery procedures with clear responsibilities. 
+

learn-pr/wwl-data-ai/ai-information-extraction/7-exercise.yml

@@ -1,12 +1,12 @@
 ### YamlMime:ModuleUnit
 uid: learn.wwl.ai-information-extraction.exercise
-title: Exercise - Extract information
+title: Exercise - Get started with information extraction in Microsoft Foundry
 metadata:
-  title: Exercise - Extract information
-  description: Use Azure AI to extract information from content
-  author: graememalcolm
-  ms.author: gmalc
-  ms.date: 05/11/2025
+  title: Exercise - Get started with information extraction in Microsoft Foundry
+  description: Use Microsoft Foundry to extract information from content.
+  author: sherzyang
+  ms.author: sheryang
+  ms.date: 01/19/2026
   ms.topic: unit
   ms.collection:
     - wwl-ai-copilot