Pegasystems Applies Predictive Analytics to Business Processes

It has been a busy spring in the business process management (BPM) market, with Adobe, Appian, Metastorm, Ultimus and Lombardi all introducing upgrades in recent weeks. Now it's Pegasystems' turns, announced on June 5 a Predictive Execution Optimizer (PXO) module designed to apply predictive models to business processes.

With the release of PXO, Pegasystems (Pega) says it can import industry-standard Predictive Modeling Markup Language (PMML) models directly into its SmartBPM suite and apply them to processes without requiring coding and development work. In contrast to static business rules and thresholds used at process decision points, predictive models gauge opportunity and risk on a case-by-case basis, assessing the data and history tied to specific process instances. Pega says PXO will likely be used to apply predictive models to many of the same business processes in which they're used today, gauging risk in lending and insurance scenarios, scoring customer segments in marketing campaigns, optimizing cross-sell and up-sell efforts, and identifying high-value customers.

"The difference with PXO is the speed with which these models can be executed without manual coding and the life cycle that these models experience once they're in the BPM suite," says Russell Keziere, senior director of BPM marketing. "We can manage, version, optimize and learn from these models within the context of BPM and its continuous improvement methodology."

Pega says it's the first BPM vendor to integrate predictive models developed within tools from SAS, SPSS and other vendors. Once the models are imported into the BPM suite, the models can be versioned, modified and optimized for different processes, customer types, regions and other variables.

"The PMML representation of the model is transformed into an easy-to-use rule form that can be edited and revised," explains Setrag Khoshafian, vice president of BPM technology. Pega has developed a white paper on PMML and the use predictive models within BPM that's available for download.

Alliances and partnerships are not uncommon between BI and BPM vendors, but BPM vendors seem to be developing business intelligence capabilities internally these says. In addition to Pega's PXO module, the recent Appian 5.6 upgrade includes "real-time analytics" designed to optimize by continuously monitoring process stats and adjusting thresholds to changing conditions. Metstorm BPM 7.5, released in May, offers a wizard-style interface that helps business users and process owners configure custom reports from the suite's process intelligence module. And Ultimus' Adaptive BPM Suite has a wizard that guides users through round-trip process optimization and "deep-dive" analytics for continuous process improvement. It's not uncommon to hear BPM vendors boast that they're leading the way in helping organizations to act on BI insight.