Connect to Amazon SageMaker models in Prediction Studio
Valid from Pega Version 8.4
Make the most of your custom Amazon SageMaker models in Pega Platform™ by connecting to the models in Prediction Studio. You can then run the Amazon SageMaker models as part of your decision strategies.
For more information, see Enrich your decisioning strategies with H2O and Amazon SageMaker predictive models (8.4).
Import H2O models to Prediction Studio
Valid from Pega Version 8.4
Make the most of your custom H2O models in Pega Platform™ by importing them to Prediction Studio. You can then include the H2O models in your decision strategies.
For more information, see Enrich your decision strategies with H2O and Amazon SageMaker predictive models (8.4).
Support for auditing adaptive model decisions
Valid from Pega Version 8.4
Pega Platform™ now stores all adaptive model scoring data so that you can identify the source of each decision, such as the exact model version that was used for scoring. With this feature, you can ensure that your application is auditable, transparent, and in compliance with regulatory requirements related to using adaptive models.
For more information, see Configuring the Adaptive Decision Manager service.
Create predictions in Prediction Studio
Valid from Pega Version 8.4
Predict customer behavior and business events by creating predictions. To create a prediction, you answer a series of questions about what you want to predict. For example, you can create a prediction to determine the likelihood of customer churn.
For more information, see Create predictions in just a few clicks (8.4).
New text analytics APIs for Intelligent Virtual Assistant
Valid from Pega Version 8.4
Pega Platform™ now provides text analytics APIs that separate Intelligent Virtual Assistant and text analyzers into independent modules. This solution prevents integration issues when you change your text analyzer configuration.
For more information, see Text analytics APIs.
Optimized performance of embedded decision strategies
Valid from Pega Version 8.4
The performance of embedded strategies has now been optimized so that these strategies take less time and fewer CPU and memory resources to complete. This enhancement increases the performance of cloud and on-premises deployments.
For more information, see About Strategy rules.
New method of aggregating real-time events
Valid from Pega Version 8.4
Event Strategy now uses the approximate median for calculating aggregations. Approximate median calculation replaces the existing method because it requires fewer system resources for saving or loading the event strategy state. In addition, you can now calculate aggregated values by using a new Landmark window type. The window type captures all values of specific event properties from the start of the data flow that references the event strategy.
For more information, see Adding aggregations in event strategies.
Extended language support in text analytics
Valid from Pega Version 8.4
Pega Platform™ text analytics now supports 9 more languages. You can build entity and topic detection models to analyze text in the Japanese, Russian, Turkish, Danish, Norwegian, Swedish, Polish, Croatian, and Czech languages. In addition, Pega Platform now includes a default small talk detection model for each of these languages.
For more information, see Language support for NLP and Out-of-the-box text analytics models.
Historical data extraction for adaptive models
Valid from Pega Version 8.4
Adaptive models now support the extraction of historical customer responses for offline analysis. Additionally, you can use the historical data to build a model in a machine learning service of your choice.
For more information, see Extract historical responses from adaptive models for offline analysis (8.4).
Improved reliability of Visual Business Director
Valid from Pega Version 8.4
Improvements to the reliability of Visual Business Director (VBD) eliminate the impact of temporary access loss to the VBD cluster. The improvements include the ability to continue writing data to a VBD dataset while the corresponding server node is unavailable, enhancements to single-case data flows that write data to VBD, and a reduction in the VBD cache size.
For more information, see Eliminate the impact of temporary access loss to the VBD cluster (8.4).