Setup Microsoft Foundry

In this section, you will set up the Microsoft Foundry resource and deploy your first model.

TIP

You do not need to deploy Microsoft Foundry. This is already done for you in your Spektra environment.

Steps

Log in to Azure
- Sign in to the Azure Portal.
Create an Microsoft Foundry Resource
- Navigate to the Microsoft Foundry service.
- Click Create a resource.

Enter the resource details
Fill in the form with the following values, then click Next:

Field	Value
Subscription:	Select the subscription provided for this workshop
Resource group:	Click `Create new` and give your resource group a descriptive name, e.g. `pizza_workshop-RG`
Name:	Enter a unique name, e.g. `pizza-foundry-resource-7yud`
Region:	Select West US (⚠️ Do not select another region)
Project Name:	`Pizza-Workshop`

Deploy the resource
- Click Next through the remaining steps until you reach Review + Create.
- Click Create to deploy the resource.
- Wait 1–5 minutes for the resource to finish deploying.
Open Microsoft Foundry
- Navigate to AI.Azure.com.
- You should now see the Microsoft Foundry projects linked to your subscription.
- Click on your project, e.g. Pizza-Workshop.
Deploy a base model
- In the project, go to Model + endpoints.
- Click Deploy model → Deploy base model.
- Select the model gpt-4o and click Confirm.
- Leave all other settings at their defaults and click Deploy.
This will make the model available in your project for use by your agents.
Test the model
- Once deployment is complete, click Open in Playground.
- In the chat window, type:
```
Hello world
```
- You should see a response from the gpt-4o model. 🎉