Skip to content

README.md

Latest commit

 

History

History
117 lines (77 loc) · 5.35 KB

README.md

File metadata and controls

117 lines (77 loc) · 5.35 KB
Raw

Portal Lab 0: Environment Validation

Welcome to the first portal lab! This short exercise ensures you have access to all the Azure resources and the Foundry Portal needed for the remaining labs. Everything here is done in the browser — no code required.

Expected duration: 10 min

Prerequisites:

  • An Azure account with access to the pre-created resource group

🎯 Objective

The goals for this lab are:

  • Confirm you can access the Azure portal and identify the provisioned resources.
  • Sign in to the Foundry Portal and navigate the pre-created project.
  • Verify that the three pre-deployed models are available and ready.

🧭 Context and Background

The coding challenges (Challenge 0–4) use a code-first approach with Python and .NET. These Portal Labs offer a parallel no-code track where you explore the same Contoso Tires manufacturing scenario entirely through the Foundry Portal at ai.azure.com.

The environment provisioned includes resources you'll interact with across both tracks:

Resource Purpose in Portal Labs
AI Foundry (project) Central hub — models, agents, tools, knowledge
Storage Account Holds wiki markdown files (used in Portal Lab 3 for Foundry IQ)
AI Search Powers knowledge retrieval behind Foundry IQ
Cosmos DB Factory data store (used by coding challenges; reference context for portal labs)
API Management AI Gateway (used by coding challenges)

Note

Not all resources in the resource group are used in the portal labs. That's expected — some resources are specific to the coding challenges.

✅ Tasks

Task 1: Explore Azure Resources

  1. Open portal.azure.com in your browser and sign in.
  2. Navigate to Resource groups and find the pre-created resource group for this workshop.
  3. Review the list of resources. You should see at least:
    • An Azure AI Foundry resource (formerly Azure AI hub/project)
    • A Storage Account
    • An AI Search service
    • A Cosmos DB account
    • An API Management service
    • Application Insights / Log Analytics

✅ You should see something similar to this

A resource group containing 12 resources.

Azure Resource Group

Task 2: Access the Foundry Portal

  1. Open ai.azure.com in your browser and sign in with the same Azure account.
  2. Ensure that the New Foundry portal experience is enabled — look for the New Foundry toggle in the upper-right corner and make sure it is turned on.
  3. You should land on the Foundry Portal welcome page. Take a moment to orient yourself:
    • Top navigation bar — Home, Discover, Build, Operate, Docs
    • Welcome banner — Shows your name and a Start building button
    • Project credentials — Displays your Project API key, Project endpoint, and Azure OpenAI endpoint
    • New arrivals — Highlights newly available models and deployment options
    • Coding quick start — Expandable section with getting-started guidance
  4. Verify that your pre-created project is visible on the welcome page — you should see its credentials (API key, endpoint) displayed. If no project appears, ask your coach for help.

✅ You should see something similar to this

The Foundry Portal welcome page showing your name, the Start building button, project credentials, and the top navigation bar with Home, Discover, Build, Operate, and Docs.

Foundry Portal Welcome Page

Tip

If you see a "request access" or permission error, ask your coach for help. You may need to be added to the project's access control (IAM) in the Azure portal.

Task 3: Confirm Model Deployments

  1. In the Foundry Portal, click Operate in the top navigation bar, then select Assets → Models in the left sidebar.
  2. Confirm you see three pre-deployed models:
Model Type Expected Status
gpt-4.1 Chat completion ✅ Succeeded
gpt-4o-mini Chat completion ✅ Succeeded
text-embedding-3-large Embedding ✅ Succeeded
  1. Click on one of the deployments to view its details: deployment name, model version, rate limits, and endpoint URL.

✅ You should see something similar to this

The Assets page under Operate showing three model deployments, all with state "Succeeded".

Model Deployments

🛠️ Troubleshooting and FAQ

I can't find the resource group in the Azure portal

  • Make sure you're signed in with the correct Azure account.
  • Try searching for the resource group name in the top search bar.
  • Check that your subscription filter (top-level filter bar) is not hiding the subscription.

I get a permission error in the Foundry Portal

  • You need at least Cognitive Services User or Contributor role on the AI Foundry resource.
  • Ask your coach to verify your role assignment in the Azure portal under the resource's Access control (IAM) blade.

🧠 Conclusion

You've confirmed access to:

  • The Azure resource group with all provisioned resources
  • The Foundry Portal project with no permission issues
  • Three pre-deployed models ready for use

Next: Portal Lab 1 — Models