Process Mining in Application Maintenance and Support — Part 1

I write about Process Mining not as a master but as a humble practitioner. There are many and superior articles and analysis, written by academicians and statisticians. Sometimes my reaction, as a practitioner, has been; can we have “plain English” please? I have tried to stay away from the formulae, attempting to focus on how Process Mining can assist with answers to issues in the industry.

This series is a dedication to that humble application log that needs its place in the sun. For many years these logs were banished to system administrators using “geeky tools” to analyze them. The study of logs has progressively become “easier” with process mining tools. Applications have moved higher on the scale of “ease of use”/ “low code or zero code” enabling inclusivity of Business/Process Analysts and System Architects.

It is difficult to write a single analysis that addresses Business and Technical practitioners. I have attempted to split this article so please feel free to read as many sections as you feel are relevant. This part introduces the potential areas to use Process Mining. The subsequent 2 parts explore how process mining gets to answer the questions and potential areas that we spoke off in this part.

So how can Process Mining help? I do not intend to replicate what innumerable articles have done. We have used process mining to validate conformance to standard procedures and reporting. These are some typical scenarios below, you may relate to any or all of them:

Possible scenarios

In the subsequent article, I intend to take Application Management and Support. Maintenance teams use a service desk application to manage (record/track/close) customer requests/ incidents. We will work with actual logs to see if we can find answers from application logs made available and how these can help us.

What’s in Part 2:

We review the actual logs from a Bank’s service desk application. Log filtering for relevance and size to obtain reasonable computation time. Reverse engineer the process from the given logs! And determine some bottlenecks.

Your comments are welcome.