THE LOGIC OF EXPERTS
With CLEVVA, you have an application that allows you to capture more than the knowledge of experts. You can capture the factors they consider when making decisions or taking actions, and the logic they apply to different contexts. The result is that you can very rapidly start offering targeted users intelligent decision support that can ensure they make the right decisions and take the right actions, and that you have a record to prove it.
When experts make decisions and take actions, they typically apply a three phased approach, each applying different forms of diagnostic logic.
The first phase typically involves diagnosing the root cause of the problem. This diagnosis looks at a set of data variables that relates to the context within which the problem presents itself.
Once the root cause has been clearly identified, the next phase is to then, based on all possible solutions, identify the right solution to address the specific cause, given the specific context. This diagnostic typically involves the analysis of the variables that differentiate each solution i.e. the expert will keep identifying differentiating factors until a single solution is left.
And once the right solution is identified, the expert then looks to action the outcome by applying standard procedural logic typically guided by policy decision rules embedded in each decision point. This means that, given the specific context, cause and solution, the decisions within the procedure will change.
How VAs Capture Expert Logic
CLEVVA has analysed these logic forms, and has identified specific logic mechanisms that are used in each phase. This allows you to build the decision-making ‘cogs’ separately, and then based on the specific context faced, CLEVVA ensures that the path taken by the user can adjust automatically without you having to hard wire specific routes before hand.
The result is that by building multi-dimensional logic components, CLEVVA is able to support decisions and actions that may have billions of possible variations or routes. By capturing logic that reflects specific decisions into separate VAs, you effectively build up a catalog of decisions that can then be linked to related decisions or VAs, depending on the requirement.
This means that each VA acts like an independent logic object that can operate on its own, or in relation to other decisions.
This re-usability ensures effective maintenance of ‘master’ versions of decision logic that can apply across multiple areas.
In addition, when a VA triggers another VA, the decisions and actions taken in the first VA are carried across to the next VA, ensuring that the context is taken into account and influences the decisions taken and the experience offered.
VA Development Life Cycle
End users can only see VA’s once they have been published. This makes it possible for the authors and experts to build VA’s on the production environment and then to simply publish the VA once they are comfortable that it has been tested and is ready for the end users to access it.
The act of publishing is made simple for the business authors, because all they need to do is change its status which will allow end users with the correct roles to access it.
VA’s can be improved once they have been published. CLEVVA provides built in version control functionality that automatically increments the version when a logic change needs to be made (data changes update automatically).
The new version of the VA will be in draft state until it is published as the latest available version.
End users only see the latest available version of the VA, so they will not be impacted while the new version of the VA is being updated and tested.