Skip to content

Migrate AI-05: Replace Azure OpenAI provisioning with Microsoft Foundry#124

Open
theresa-i wants to merge 4 commits into
mainfrom
update-ai-05-foundry-migration
Open

Migrate AI-05: Replace Azure OpenAI provisioning with Microsoft Foundry#124
theresa-i wants to merge 4 commits into
mainfrom
update-ai-05-foundry-migration

Conversation

@theresa-i

Copy link
Copy Markdown
Contributor

Summary

This PR updates the AI-05 (Evaluate Models) lab to use Microsoft Foundry for resource provisioning and model deployment, replacing the previous Azure OpenAI service flow.

Changes

  • Replaced "Provision an Azure OpenAI resource" section with "Create a Microsoft Foundry resource and project" — uses the new Foundry resource creation portal link and adds steps to create a project within the resource
  • Added Cloud Shell step to obtain a short-lived Entra ID access token via az account get-access-token --resource https://cognitiveservices.azure.com (replaces API key auth)
  • Updated model deployment to use the Foundry deployments page, targeting gpt-4.1 (version 2025-04-14)
  • Narrowed supported regions to North Central US and Sweden Central (quota-safe for gpt-4.1)
  • Updated primarytopics metadata: replaced Azure OpenAI with Microsoft Foundry
  • All Databricks cluster, notebook, MLflow evaluation, and prompt-improvement content is unchanged

Review notes

Please verify the Foundry resource creation steps and token auth flow match the current portal experience. The lab hoster (Skillable) should confirm the provisioning steps work end-to-end in a hosted lab environment.

Changes from main:
- Updated lab title and description to reference Microsoft Foundry
- Replaced 'Provision an Azure OpenAI resource' section with 'Create a
  Microsoft Foundry resource and project', using the new Foundry resource
  creation portal link and Entra ID (access token) auth instead of API keys
- Added step to create a project within the Foundry resource after deployment
- Added Cloud Shell step to obtain a short-lived access token via
  `az account get-access-token --resource https://cognitiveservices.azure.com`
- Narrowed region list to North Central US and Sweden Central (quota-safe)
- Updated model deployment target from Azure OpenAI service to Microsoft
  Foundry deployments page, deploying gpt-4.1 (model version 2025-04-14)
- Updated primarytopics metadata: replaced 'Azure OpenAI' with
  'Microsoft Foundry'
- Preserved all Databricks cluster, notebook, MLflow evaluation, and
  prompt-improvement content unchanged"
@SkillableKarl

Copy link
Copy Markdown
Contributor

I think in this .md, in Task 'Evaluate LLM with a custom function', Step 2, there is a typo: I am not certain if the intent is to have student use Copilot/Genie Code to fix this on their own, or - more likely - its just a typo. In any case, dropping here for reference:

image

@theresa-i

Copy link
Copy Markdown
Contributor Author

Thanks for the sharp eyes, @SkillableKarl! 🙏

The "appropiate""appropriate" typo and the unescaped inner quotes in the template string were already fixed earlier in this PR. I've just pushed one more small cleanup (commit 70d8241):

  • api_version updated from "2024-08-01-preview" to "2025-04-01-preview" to match AI-03, AI-06, and AI-07
  • Template string in Step 2's eval dataset now uses a single-quoted outer string ('...') instead of backslash-escaped double quotes — cleaner and less confusing for learners copying the code

Thanks again for reviewing and catching that — really valuable having Skillable eyes on these labs before they go live!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants