Segmentation techniques were deployed to predict the likelihood of an IT Ticket getting reassigned in an online IT service management application. A robust segmentation model was built, despite the predictor variables being largely categorical in nature. Custom logic was developed to circumvent the absence of an in-built R function to implement the CHAID algorithm. Distinct segments of IT incidents were identified that were prone to reassignment to various degrees, providing the client a handle to exercise greater care in initial assignment of those IT incidents that were more prone to reassignment.
Read MoreA telecommunications services provider wants to allocate its service engineers to various customer locations in a manner that minimizes engineers’ travel requirements while still ensuring customer satisfaction under its service level agreements (SLAs). Segmentation techniques are deployed for the various exchanges across which the company operates. Optimization techniques are put in place to develop a robust work allocation framework that meets demand-side requirements amidst supply-side constraints in a fashion that minimizes engineers’ travel requirements. The optimization model is evaluated across various scenarios that model peaks & troughs in service demand as well as engineers’ availability.
Read MorePredictive analytics techniques were deployed to predict the frequency of an IT Ticket getting reassigned in an online IT service management application. A robust predictive regression model was built, despite the predictor variables being largely categorical in nature. Poisson regression techniques were deployed to circumvent the non-normal behaviour exhibited by the target variable. The regression model enabled effective workforce planning to resolve IT incidents, based on the predicted reassignment frequency of the associated IT Ticket. The model also provided early warning signals of likely multiple reassignments, allowing better proactive management of high-priority IT incidents.
Read MoreA telecommunications services provider wants to identify media channels wherein its advertising campaigns yield the maximum utility in terms of revenue generation. Predictive analytics techniques are deployed to model the relationship between revenues and advertisement spending. Separate models are developed for the different media channels, viz, television, radio, newspapers and billboards. The lagged effect of advertising on revenues is accommodated, and variable transformation techniques are used to convert the inherent non-linear relationship between revenues and advertising into an approximately linear one. The resultant marketing mix modelling enables the company to identify the media channels that are most effective in generating revenues, together with the optimal levels of advertisement spending beyond which the beneficial effects of further advertising saturate with no further meaningful incremental revenue generation.
Read More