Skip to main content


         This documentation site is for previous versions. Visit our new documentation site for current releases.      
 

Activating and training adaptive models for artificial intelligence

Updated on December 21, 2020

Artificial intelligence in Pega Sales Automation helps you to proactively assess risks on deals in the pipeline, coach newly recruited sales representatives, and identify leads that have a high probability to be converted to opportunities. Before using artificial intelligence insights with Pega Sales Automation, activate the feature for your implementation and then configure the application to train Pega’s adaptive models for artificial intelligence.

To configure your application for artificial intelligence, log in to Pega Sales Automation and complete the following steps:
Note: If you installed the Pega Sales Automation sample application, as an operator with the Sales Ops persona, reset the artificial intelligence sample data to the original form by using the Tools menu.
  1. In the navigation pane of App Studio, click SettingsApplication Settings.
  2. On the Features tab, select the Artificial intelligence insights - opportunity insights, lead ranking, and sales coach check box to see all of the AI capabilities.
  3. Enable any of the listed AI capabilities by selecting the individual check box.
  4. Click Save.
  5. Switch to Dev Studio.
  6. In the navigation pane of Dev Studio, click RecordsSysAdminAgent Schedule.
  7. Search for an open the SA-Artifacts agent schedule.
  8. On the Edit Agent Schedule page, select the Enable this agent check box.
  9. Click Save.
  1. In the header of Dev Studio, click ConfigureDecisioningInfrastructureServices.
  2. Verify that each of the following services on the list contains a node with a status of Normal:
    • Decision Data Store
    • Adaptive Decision Manager
    • Data Flow
    • Visual Business Director
  1. In the header of Dev Studio, click ConfigureApplicationDistributionImport.
  2. Click Choose File, browse for and select the HistoricalData file from your distribution media, and then follow the wizard instructions.

    Pega-provided historical data consists of a snapshot of data from a production environment for various models.

  1. In the navigation pane of Dev Studio, click RecordsData ModelData Set.
  2. Search for and open the pxDecisionResults data set of the Data-Decision-Results class.
  3. Click ActionsRun.
  4. In the Operation field, select Truncate.
  5. Click Run.
  6. Repeat steps 2 to 5 for the PreviousStages data set of the SA-SR class.
  1. In the header of Dev Studio, cllick ConfigureDecisioningModel Management.
  2. Select all of the models listed below:
    • PredictWin
    • PredictMoveNextStage
    • PredictCloseDate
    • BaseWinModel
    • LeadRanking
    • PredictEffectiveness
  3. Click Delete Models.
Run the data flows for opportunity insights to pass the incoming data to the adaptive models so that the system can calculate opportunity insights.

Note: Before performing the steps below, make sure that you have the EnableOpportunityInsights dynamic system setting set to true.

  1. In the navigation pane of Dev Studio, click RecordsData ModelData Flow.
  2. Search for and open the StoreOpportunitySnapshots data flow.
  3. Click ActionsRun.
  4. After you see the confirmation message with a successful result, repeat steps 2 and 3 for the TrainFromHistory data flow.
Run the data flows for the sales coach to pass the incoming data to the adaptive models so that the system can calculate sales coach suggestions.

Note: Before performing the steps below, make sure that you have the EnableSalesCoach dynamic system setting set to true.

  1. In the navigation pane of Dev Studio, click RecordsData ModelData Flow.
  2. Search for and open the StoreSalesRepSnapshots data flow.
  3. Click ActionsRun.
  4. After you see the confirmation message with a successful result, repeat steps 2 and 3 for the CaptureEffectivenessOutcomes data flow.
Run the data flows for lead ranking to pass the incoming data to the adaptive models so that the system can calculate the B2B lead scores.
  1. In the navigation pane of Dev Studio, click RecordsData ModelData Flow.
  2. Search for and open the StoreLeadSnapshots data flow.
  3. Click ActionsRun.
  4. Repeat steps 2 and 3 for the CaptureLeadOutcomes data flow.
  5. Optional: To set up predictors and calculate lead score for already existing leads and store them in the lead predictor table, run the InitialiseLeadPredictorTable activity.
Run the data flows for lead ranking to pass the incoming data to the adaptive models so that the system can calculate the B2C lead scores.
  1. In the navigation pane of Dev Studio, click RecordsData ModelData Flow.
  2. Search for and open the StoreIndividualLeadSnapshots data flow.
  3. Click ActionsRun.
  4. Repeat steps 2 and 3 for the CaptureIndividualLeadOutcomes data flow.
  5. Optional: To set up predictors and calculate lead score for already existing leads and store them in the lead predictor table, run the InitialiseLeadPredictorTable activity.
  1. In the navigation pane of Dev Studio, click RecordsSysAdminDynamic System Settings.
  2. Search for and open the usePreloadedNBA dynamic system setting.
  3. In the Value field, set the value true.
  4. In the navigation pane of Dev Studio, click RecordsSysAdminJob Scheduler.
  5. Search for and open the GenerateNBA job scheduler.
  6. Turn on the Enable Job Scheduler switch.
    Tip: To see the job scheduler changes instantly, run the LoadNBAForAllOpps data flow.
  7. Optional: If your implementation layer has additional next best actions, add them by overriding the LoadNBAForAllOpps_Ext data flow. Then, create declare triggers to track work item updates.

Have a question? Get answers now.

Visit the Support Center to ask questions, engage in discussions, share ideas, and help others.

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us