As part of its efforts to drive greater insight into customer dynamics via predictive analytics, IBM has developed a series of integrated predictive analytics solutions, one of which is called Predictive Customer Intelligence (PCI). Zementis offers a Functional Accelerator that is integrated with IBM’s PCI solution.
IBM Predictive Customer Intelligence offers businesses critical information and insights that enable them to provide proactive service to their customers, develop a consistent customer contact strategy and improve their relationships with customers. The technology integrates information from multiple internal and external data sources, enabling holistic analysis that yields predictive insights. These insights, in turn, allow businesses to make informed decisions that actively shape and improve the customer experience while increasing customer lifetime value (CLV).
A key component of this solution is a library of accelerators that IBM makes available via IBM’s Analytics Zone, its portal for analytics solutions. Visitors to the Analytics Zone will find a variety of accelerators developed for specific industries and cross-industry / horizontal use cases (“functional accelerators”). A description of IBM”s Predictive Customer Intelligence offering and the various accelerators is accessible here, and a technical brief covering the integrated Zementis / IBM solution for predictive analytics is available here.
The Zementis / IBM functional accelerator includes an example that simulates a customer’s interactions through a channel, such as a call center. The call center retrieves a customer’s profile and evaluates it first for the propensity to churn and then for overall sentiment. The churn score from the first model is then used as an input for the second model, the sentiment analysis.
The functional accelerator demonstrates how businesses can use models from two different modeling tools in a real-time scenario. It combines the real-time scoring capability of the Zementis ADAPA (Adaptive Decision and Predictive Analytics) engine with the modeling of IBM SPSS. Users can extend the library of available analytical and predictive models to include any models that can be published in Predictive Model Markup Language (PMML).
For more details or if you have any questions, please feel free to contact us!