Virtually predictive

I was at a dinner party recently where one of the guests was trying to explain the concept of virtualization applications, in the context of a start-up he was trying to get off the ground. It started me thinking that the intersection of analytic applications with virtualization have the potential to provide a much higher level of strategic insight across the enterprise, particularly in complex, collaborative environments. Several years ago I worked with a predictive analytic start-up that was focused on IT Operations planning for manufacturers; our intent was to predict subsystem failure before the event itself. The application was remarkably accurate, but we never really took it past predicting failure rates for very specific areas. Integrating a predictive engine with a virtualization engine could have allowed us to run simulations across the entire information and production lifecycle management process. The main reason we didn’t develop this further was that at that time virtualization technology was still in its infancy, but that is no longer the case. Rather than saying “what if this specific thing happens”, we could now say “what if we change the whole model”, or “what is the highest probability effect of changing out our suppliers, or outsourcing a critical process”? This is another potential technology confluence point coming up, and one that is well worth keeping an eye on. More on this later.