Unable to parse the response from the Azure ML Web Service in Azure Stream Analytics

Question

Hi, using Azure ML Studio I have created an endpoint for a model generated with automated ML. The model works fine in test (consume) - provides an expected outcome. Then I created a Stream Analytics query using the function to consume the same ML endpoint. However when I test the Stream Analytics query I receive the following error:

"Callout failed within query runner. An error was encountered while calling the Azure ML web service. An error occurred when parsing the Azure ML web service response. Please check your Azure ML web service and data model. The content of the response from the ML web service should be a JSON array. The response received from the Web Service is: {"Results": ["none", "none", "none"]} Parameter name: result"

The result I am getting is fine - as expected - but the problem seems to be with parsing the result.
So by reading the docs I understand the desired output format from ML endpoint is JSON Array like this ["none", "none", "none"], while I am getting a JSON object. {"Results": ["none", "none", "none"]}
The question is, can I (how I) modify the output format (swagger?) to have it return the json array? The model was calculated by automated ML (no-code)

For the record the Stream Analytics query is like this:

SELECT udf.pdmpredict(TRY_CAST(inputArray AS record))
INTO [pdm-predict-data]
FROM ModelInput
WHERE inputArray is not null

udf.pdmpredict is my ASA ML function created in accordance to this article: https://learn.microsoft.com/en-us/azure/stream-analytics/machine-learning-udf

Any ideas will be greatly appreciated!
Thanks.