AIStudioDeployment.md

AI Studio Deployment Guide

Please follow the steps below to set up the AI Studio Hub, Project, PromptFlow and Deploy as endpoint. Clone or download this repository before you start following the steps below.

Step 1: Create AI Studio Project

1. Launch the AI Studio and select Build from the top menu.

   

2. Click on New AI Project.

   

3. Enter a Project Name and click on Create a New Resource.

4. Enter a new AI Hub Resource Name. And then select Azure Subscription, Resource Group, Location from drop-down menus. You should choose the same resource group you created from ARM deployment in previous step.

   IMPORTANT: Do not create new resources in this step and link to the ones created during one-click deployment. This applies to Resource group, Azure OpenAI and Azure AI Search fields.

   

   

5. Select Azure OpenAI resource from drop-down menu.

6. Select Azure AI Search resource from drop-down menu.

7. Click on Next button.

8. Review the details and click on Create an AI Project button.

   

Learn more about AI Studio Projects

Step 2: Import Prompt Flow and Deploy

1. Click on PromptFlow button from left menu under Tools.

   

2. Click on Create button. Then click on Upload button from Upload From Local option from the last row.

   

3. Click on Zip File radio button. Then click on Browse to select the file DraftFlow.zip from the cloned/downloaded GitHub repository folder. The file will be located at <Your Download Folder Path>/Deployment/scripts/ai_hub_scripts/flows/.

   

4. Once the DraftFlow.zip file is uploaded, change folder name to DraftFlow and Select Flow type as Chat Flow and click on Upload button.

   

5. Click on Select runtime and chick on Start from the drop-down list. It can take few minutes for the runtime to start.

   

6. Click on Deploy button once it is enabled. Enter a unique name for Endpoint Name field.

   IMPORTANT: This name has to be unique across all endpoints in your Azure subscription.

   You can leave the Deployment name, Virtual machine type, and Instace count as populated and click on Review + Create. Optionally you can choose a different VM type/size and increase/decrease the Instance count as needed. Then review details and click on Create in the next screen.

   

7. It will take few minutes for the flow to be validated and deployed. Click on Deployments from left menu. You might only see the Default_AzureOpenAI deployments in the page until the deployment is completed. Please wait and click on Refresh after few minutes.

   

8. Click on the deployed endpoint with name draftsinference-1.

9. Click on Consume from the top menu. Copy below details to use later in step 3.6.

• Deployment

• REST endpoint

• Primary key

  

Step 3: Update the deployment keys in Azure App Service configuration

1. Launch the Azure Portal Azure Portal.

2. Enter Resource Groups in the top search bar.

   

3. Locate your Resource Group you selected/created during one-click deployment and click on it.

4. Locate the App Service in the Resource Group and click on it.

5. Click on Environment Variables from left menu under Settings.

   

6. Modify the below variables with values collected in step 2.9 above.

• AI_STUDIO_DRAFT_FLOW_ENDPOINT
• AI_STUDIO_DRAFT_FLOW_API_KEY
• AI_STUDIO_DRAFT_FLOW_DEPLOYMENT_NAME

7. Click on Apply button at the bottom of the screen. Then click on Confirm in the pop-up.

   

8. Click on Overview from the left menu. Then click on Restart button in the top menu. Then click on Yes in the pop-up message.

   

Step 4: Add Authentication in Azure App Service configuration

1. Click on Authentication from left menu.

2. Click on + Add Provider to see a list of identity providers.

3. Click on + Add Provider to see a list of identity providers.

4. Select the first option Microsoft Entra Id from the drop-down list.

5. Accept the default values and click on Add button to go back to the previous page with the identify provider added.