Analytical Process Controlling

In my professional career over the last 20 or so years, there were three major areas that I had interest in: applications, integration, and analytics. When those three are coming together and you intersect the respective capabilities, there is a new concept possible that would allow organizations to get insight into the operational truths about their business processes. Of course, there are things like Business Activity Monitoring (BAM) and Business Process Management (BPM), but none of those tools or platforms are able to provide the level of information required to find out

• Why is order management taking twice as long on Tuesdays?
  
• Why do we receive more complaints with customers in Frankfurt?
  
• Why are invoices to gold customers never sent out on time?
  

I'm working on a concept that I call Analytical Process Controlling (APC). With APC, process execution details are extracted from every business process and their underlying applications, stored in a specific data warehouse and later analyzed with data mining tools, such as correlation engines, time series and neural network algorithms. The basis for delivery of the execution steps are time-stamped Process Instance Identifiers (PIID) and agents that need to be embedded in between process steps (before and after) and also within the applications that are called from the process. All agents send the appropriate information to a listener, that forwards the PIID data to the data warehouse.

I discussed this with a few people in the industry and feedback was mostly positive. Interested in your opinion ....