Post thumbnail

Developing a product centered on analysing process performance and driving operational improvement, we have come to realize that taking the shortest-path possible is always the right solution to increasing speed and efficiency. In many use-cases, however, this path is not clear from the start, and powerful business intelligence is needed to discover it.

In incident management, issues get addressed quickly only when diagnosed correctly, as any mistakes in the routing result in a dramatic increase in resolution time. And so, deciding how to route incident tickets becomes the highest responsibility in incident management. This inspired our decision to expand the functionality of our analytics solution to cover the need for routing that is truly smart.

Why we built this?


uploadThe way software solutions currently on the market support the routing challenge is through the introduction of automation rules. By creating a rigid structure of conditions, newly raised incidents will be routed based on these business rules – following the organization’s idea of “the shortest path”.

However, if these automation rules are not correct 100% of the time and account for 100% of the possible situations, then the potential for routing is not fulfilled.

This is when we decided to look into how currently available technologies can eliminate both the need for configuration of business rules, as well as the human factor in ticket routing. We are today announcing the first version of our solution to this challenge – Smart Routing.

What makes the Cupenya routing so smart?

The Cupenya approach looks to skip the configuration of such rules and instead learn from the patterns found in ticket history. We do this by creating a comprehensive classification of incident tickets based on all of their structured (and soon also unstructured) parameters. We enrich this classification by extracting the records of the last action taken before the successful closure of a ticket, and then retrace the steps that preceded it, effectively constructing a template for the shortest path solution of each incident type.


upload-1Where this system truly shines is in it’s ability to have ongoing learning and re-classification, adapting itself to new business requirements, changes in procedure, and the introduction of new services and products. This means that the routing will not need to be reconfigured, with new rules manually introduced to handle every exception. Instead, any changes will be recognized automatically, classified and stored for future use by the Cupenya machine learning algorithm.

What’s next?

Smart routing, matched together with performance-triggered Action Cards takes a giant step forward towards truly prescriptive operational analysis. We are now aiming at iteratively improve this offering, adding a capability to feed unstructured data to the machine learning algorithm that handles the smart routing. Adding this cognitive analytics aspect to the solution is the next step towards creating the comprehensive operational process improvement suite.

And as an added bonus, we are soon unveiling a whole new view of Cupenya’s interface, where users can track the week-over-week improvement driven by the use of Action Cards. We are calling it Impact Center, and will make it available to all users in an upcoming update.