Using AI to make business cases easier

One of the elements of my job as Technical Director is to provide tools that help businesses make good decisions about their future operations. These decisions are based on a multitude of factors, but the one factor I am interested in is how these decisions can impact the company’s future financial performance. This involves analysing the options a business has, understanding the processes affected by those options and then quantifying, in financial terms, the impact those options may have on the business, in other words a business case.

It is the individual nature of a business that limits the value of using benchmarks in business cases

One thing I think about a lot is how to make this process as easy and seamless as possible. Obviously, software interface design is a big factor in this, but my real interest is in the procedural and technological methods we can leverage. Another factor that is often mentioned is benchmarks, but for the very reasons mentioned above I find these have very limited use in creating business cases that are meaningful and specifically reflect a particular or unique business position. It is the individual nature of a business that limits the value of using benchmarks in business cases. That’s not to say that they should not be used! On the contrary, to market or articulate the potential impact something may have on a business then they can be very valuable.

My most recent interest has been around the development of artificial intelligence and most recently we have now integrated an element of analytics or self learning into our software. Each time a user completes some work, the way in which they have used the software and the combinations of options are stored. From this relationship, strengths are then calculated between the various elements of the software, enabling the software to learn what combinations are most used. This in turn allows the software to suggest to the user a possible future pattern of use based upon its own “experience”. These suggestions are aimed at helping the user make choices based upon the software’s experience, but is this intelligence? This is really our first - but very exciting - toe in the water of implementing such learning inside software but it is an area we intend to expand on a great deal.

We are also working feverishly on the re-engineering of our flagship product that will enable lots of other very exciting things, but more on that to follow in due course!