Connecting Cognitive Services using Azure Logic Apps for AI Insights

Question

Pulumi · Accepted Answer

To connect Azure Cognitive Services with Azure Logic Apps for AI insights, we will be creating an example that demonstrates how to provision necessary Azure resources using Pulumi. We'll focus on setting up an Azure Cognitive Services Account, which will provide various AI services and an Azure Logic App that will consume these AI services to generate insights.

Here's what we are going to do:
1. Provision an Azure Cognitive Services Account: This will give us access to AI services such as Vision, Speech, Language, Decision, and Search.
2. Set up an Azure Logic App: Logic Apps help us automate workflows without writing code. We will use it to integrate with Cognitive Services and process the AI insights.
3. Configure a simple Logic App workflow: Although defining the entire workflow is beyond the scope of this code, we will create a placeholder for where you would use the Cognitive Services.

Below is the Pulumi program written in Python that creates these resources:

```python
import pulumi
import pulumi_azure_native as azure_native

# Configurations for our resources
resource_group_name = 'ai_insights_group'
cognitive_services_account_name = 'ai_cognitive_account'
location = 'West US'  # Replace with the location you want to deploy the resources in

# Create an Azure Resource Group to organize all the resources
resource_group = azure_native.resources.ResourceGroup(
    resource_group_name,
    location=location
)

# Create an Azure Cognitive Services Account
cognitive_services_account = azure_native.cognitiveservices.Account(
    cognitive_services_account_name,
    resource_group_name=resource_group.name,
    location=resource_group.location,
    sku=azure_native.cognitiveservices.SkuArgs(
        name="S0"  # The pricing tier. S0 is standard tier, you may choose the tier based on your need
    ),
    kind="TextAnalytics",  # Specify the kind of Cognitive Service you want to create
    tags={
        "Environment": "Dev"
    }
)

# Create an Azure Logic App
logic_app_name = 'ai_logic_app'
logic_app = azure_native.logic.Workflow(
    logic_app_name,
    resource_group_name=resource_group.name,
    location=resource_group.location,
    definition={
        "$schema": "https://schema.management.azure.com/providers/Microsoft.Logic/schemas/2016-06-01/workflowdefinition.json#",
        "actions": {
            # Define your Logic App actions here. This is a place holder for connecting Cognitive Services
        },
        "connectors": {
            # Define your Logic App connectors here. This is a place holder for connecting Cognitive Services
        },
        "triggers": {
            # Define your Logic App triggers here. This is a place holder for connecting Cognitive Services
        },
        "contentVersion": "1.0.0.0",
        "outputs": {},
        "parameters": {},
        "triggers": {}
    },
    tags={
        "Environment": "Dev"
    }
)

# Export the primary endpoint key of the Cognitive Services account which can be used to authenticate API calls
primary_endpoint_key = pulumi.Output.all(resource_group.name, cognitive_services_account.name).apply(
    lambda args: azure_native.cognitiveservices.list_account_keys_output(
        resource_group_name=args[0],
        account_name=args[1]
    ).apply(lambda account_keys: account_keys.primary_key)
)

pulumi.export('primary_endpoint_key', primary_endpoint_key)
pulumi.export('logic_app_id', logic_app.id)
```

In this program:

- We define a resource group named `ai_insights_group` to contain our Cognitive Services account and Logic App.
- We create a Cognitive Services account of kind `TextAnalytics`; this can be tailored to the specific AI service you need.
- The Logic App, named `ai_logic_app`, is defined with placeholders for actions, connectors, and triggers that will be required to interact with the Cognitive Services account.
- Finally, we export the primary endpoint key of the Cognitive Services account, which will be necessary to make authenticated requests to the AI services.

This program is just a starting point to set up the foundation for connecting Azure Cognitive Services with Azure Logic Apps. The actual workflow definitions in the logic apps, which would include specific cognitive service actions and data transformations, would depend on the specific insights you are trying to derive. Once the Logic App definition is in place, it will perform actions based on triggers and use the Cognitive Services account to process and analyze data as needed.