Using Post Method - Decision Model

Send the inputs to the ML model by uploading a JSON file corresponding to the selected ML model.

1. Either drag and drop or browse the required JSON file.

2. Click SUBMIT. The Simplified view of the list of input fields to be sent to the selected ML model appears.

3. Click Parameter Value field of any of the input to modify the required value. By default, all the input fields are selected. You can clear the check box beside the required Parameter Value if you do not want to send the respective input field to the selected ML model. Alternatively, click Standard tab to view and modify the inputs using standard JSON editing method.

4. In the Model Response list, select the parameter to store the JSON response.

5. Click Advanced Settings to update the request method, request header and to capture the response header if required. The Advanced Settings dialog box appears. The Input Format displays the format of sending the input to the ML model. The URL field displays the ML model REST API URL.

6. In the Request Method list, select GET if you want to change the request method.

7. The Accept and Content-Type header details that are added in Headers section while configuring the ML server in the Admin menu of Automation Studio auto populates in the Model Headers screen, as shown in the below screen shot. This lets you inherit the information related to the ML model REST API request and response. By default, JSON format is supported if the headers are not defined.

8. You can manually add headers by clicking the (Add) icon as per your requirement.

a. In the Header Name field, enter the name of the header.

b. In the Header Value field, select the relevant parameter holding the value of the specified header. You must predefine the parameter and the respective value in the Parameter bar to use this option. Alternatively, select the check box available beside the Header Value field and enter the header value that you want to set as the default value.

c. Repeat step x to step xii to add multiple headers.

d. You can use the X icon to delete the added headers.

9. In the Response Header list, select the parameter to store the header response.

10. Click SAVE.

11. Close the Model Inputs and Response (POST) dialog box.

The Post method of retrieving the predictions is configured. The JSON activity can then be used to retrieve the required output fields from the ML model response JSON and configure the automation process flow accordingly.