Execute Custom ML Endpoint

The code snippet given below allows you to pass input data to a published QuickML endpoint and retrieve the inference results from the ML model. The response contains the prediction generated by the ML model based on the input features provided.

The ZCQuickML instance is initialized using the getInstance() method. This initialization does not make a server-side call.

You must then create a HashMap containing the input data required by the ML model as key-value pairs. The keys in the HashMap must match the feature names expected by the trained model, and the corresponding values must contain the input data for those features.

The endpointKey is the unique identifier of the endpoint published for the ML model in your Catalyst project.

The endpoint key and input data are passed to the runInference(endpointKey, input_data) method to execute the ML endpoint. The runInference() method returns a ZCQuickMLDetail object containing the inference response.

Sample Code Snippet

//ML-endpoint change method name for ml endpoint predict to runInference
HashMap input_data = new HashMap<>();
// Give the column names and values based on your dataset.
map.put("<FEATURE_1>", “<VALUE_1>”);
map.put("<FEATURE_2>", “<VALUE_2>”);
map.put("<FEATURE_3>", “<VALUE_3>”);
ZCQuickML quickMlInstance = ZCQuickML.getInstance();
String endpointKey = “<ENDPOINT_KEY>”;
ZCQuickMLDetail result = quickMlInstance.runInference(endpointKey, input_data);
System.out.println(result.getResponse());

The syntax of the model response received is shown below:

{
  "status": "success",
  "result": [
    {
      "prediction": "1",
      "confidence": 0.87
    }
  ]
}

Parameters Used

Where to Find the Endpoint Information?

Create an endpoint for your trained ML model and open the endpoint details page in the Catalyst console. The endpoint details page provides information such as the Endpoint URL, required headers, and sample request and response .