Example

This example demonstrates how to use this AI & ML Agent to ask an Azure OpenAI model instance the answer to a math problem: to decuple the incoming sensor value.

Refer to configuration to understand all configuration options of this Agent.

Step 1: Add the Agent

Drag the Azure OpenAI Action Agent onto the canvas.
Rename the Agent, link the input endpoint to the sensor data and the output endpoint to the printer
Save the Data Stream.
Double-click to configure the Agent.

Step 2: Configure Connection

Enter the Endpoint, Key Credential, and Deployment Name. In this case, keep use variables ticked and select the variables that represent the connection values.

Step 3: Configure Messages

Click the plus button to add one or more messages, making sure to click Apply to add the message to the list. Each message is made up of a description, order, type, and message text.

In this example, add the following three messages:

A system message with the description "Assign Calculator Role" and the message text "You are an XMPro Calculation agent, designed to enrich event data in a Data Stream." This message informs the AI model of what role to assume.
A system message with the description "Format Response" and the message text "Your response must be limited to the result of the calculation only." This message informs the AI model how to format the response.
A user message with the description "User Multiply x 10" and the message text "What is {{Value}} multiplied by 10?" This message asks the model a question using a value from the input payload.