We recently participated in a process mining challenge. We received logs and transactional information from a support and problem management system called VINST, performed detailed analysis of various efficiency and performance factors and identified some key actionable patterns for improvement. It was an interesting exercise, but why are we discussing it here?
Please refer my previous post – “Leveraging Process Mining”. A service oriented approach to hybrid cloud architectures will throw open interesting approaches to improve operational efficiency. A key component of this will be the ability to drive efficiencies in re-designing human dependent processes leveraging approaches such as process mining. A whole new category of applications called smart-process apps are being discussed in this space and we shall revisit this topic in a separate post.
This analysis is focused on incident and problem management.
What did we discover? Among other things , a couple of aspects( highly applicable to every IT environment)
a) Service Request’s – tickets – were being logged and/ or assigned directly at various levels , resulting in ‘Ping-Pong’ behavior, delayed resolution times and a phenomenal backlog. We noticed this pattern when we pivoted transaction logs against the support level. We expected to see the pre-dominant trend as : level 1 -> level 2 -> Level 3. But, look what we found:-
b) Case state transition was all over the place. We created an expected state transition model ( please see below) based on understood case flow behavior, but the actual state transitions were all over the place and lacked a consistent method.Upon closer scrutiny we understood that this was the result of flawed set-ups and that over a period this resulted in abuse of case statuses/ sub-statuses.
For the current context we did not dwell into organization and people re-design( for example, multi-skilled resources can handles issues thrown at them at any level without passing them on) , but the issue of figuring out efficiency improvement points within the current people/skill structure.
To get all the insights we uncovered , please click here-> http://ceur-ws.org/Vol-1052/
Let us know your thoughts.