Many companies have built out their data infrastructure, composed of large data warehouses, varying data marts, integration processes, quality controls and much more. And, it has driven tremendous value in the form of stronger business intelligence, better views of the customer, faster business processes and more. But, with all the help on the analysis side, where’s the evolution in putting those findings into operational practice. Like quickly identifying that a premium/platinum-level client was put on hold by a customer care agent for more than 5 minutes (and no, the “we are experiencing unusually high call volume” does not let anyone off the hook).
InformationWeek’s May 24, 2010 cover story was titled “Are your Apps Smart Enough” and it prominently featured a BI report card. Here were the grades:
Real-Time insight: F
This looked like one of my middle school report cards that I either tried to doctor or just hide from mom.
Yes, an F for real-time insight, aka ‘operational intelligence.’ But, I won’t oversimplify the significant technological and organizational challenges around real-time insight. A slew of technologies and ideas have been introduced that in one way or another promised to shorten the time between incident, notification and resolution. Some vendors go a step further in jumping directly over the idea of real-time insight and evangelize a concept of predictive insight. Although nice sounding, we rarely witness a runner that cannot walk, and a singer that cannot talk (if you can tell me who I ripped that analogy off of, I’ll buy you a cookie).
The truth of the matter is that in order to gain real-time insight there has to be the capacity to tap into a number of sources of information and then identify the salient pieces of data (that may come from simple issues or complex disparate ones) that should be bubbled up. And, there’s an evolutionary part in that as we build out identifiers more become obvious and we tune, refine and often complicate our rule sets. The enabling technology most appropriate for doing this is called “Complex Event Processing” (CEP). Further and perhaps most important is that these endeavors require a very strong engagement between IT and business lines and thus strengthen that tie. In an ideal scenario, IT helps to establish connectivity to those sources of events/information and the business articulates the logic/scenarios to watch.
Operational Intelligence is the Benefit, CEP is the enabler
I tend to think “operational intelligence” sounds like a pragmatic label for real-time intelligence, or business activity monitoring. So, operational intelligence is the benefit, and CEP is a key enabling technology. Again, in general terms we’re talking about identifying and correlating a series of activities that can indicate threats and opportunities. Sometimes it’s easy to spot these and other times there’s a complex set of circumstances that when seen together or composited will be indicative of an issue or opportunity. So, what are some areas to leverage CEP to gain operational intelligence? Fraud, compliance, quality assurance, homeland defense, utility optimization?
Next post – 3 examples of CEP uses to improve operational intelligence.