Salesforce Docs Portal

Enable the Propensity to Escalate Insight Component

For the Propensity to Escalate model to work as expected, you must modify the associated batch data transformation to integrate the Einstein predictive model.

Required Editions

  1. From the app launcher, go to Data Cloud > Einstein Studio.
  2. Open the Predicted is_escalated__c Einstein studio model, and click Activate.
  3. Go to Data Transforms, and open the SI Scoring DMO. Then select Edit from the dropdown menu. The mapping for the batch data transform displays.
  4. To add the AI model node, you must edit the SI Scoring DMO batch data transformation.
    1. Click the Rename transform node and navigate to the bottom edit panel.
    2. From the edit panel, scroll down to the last transformation Dummy isEscalated Prediction, and delete it.
    3. Next, delete the Output transformation. You’ll create a new output node in later steps.
    4. Add a AI model after the Rename transformation.
    5. From the model names, select the predictive model you created earlier-- Predicted is_escalated__c, and click Next. In the Map Fields tab, map the appropriate input fields (#Case Age, Case Type, Channel, #EscalatedCases, Priority, CalculatedQueue). (Optional) In the Modify Output tab, add the number of top predictors needed in the model prediction. Click Apply .
    6. From the AI model, add an output node. Name the output DMO, and select Case ID as your primary key, and Key Qualifier Case as your FQK.
    7. Scroll down the new DMO fields, and make sure the following prediction-based fields are present: #Predicted Escalated, Predictor 1, Predictor 1 Impact, Predictor 2, Predictor 2 Impact. Click Apply.

After you’ve completed these modifications, the batch data transformation integrates the Einstein predictive model to generate predictions.