With their bespoke AI learning tool Requiment, Glasgow-based Pulsion Technology is turning the tables on the current high fail rate of many digital tranformation projects


ARTIFICIAL intelligence (AI) is increasingly becoming a highly effective way for businesses to improve their business processes. However, the successful implementation and integration of AI and machine learning into a business’s day-to-day operations is still very much a specialist discipline. 

John McGuire, founder and managing director of AI and machine learning specialists, Pulsion Technology, points out that a proper scoping and specification of a proposed new system is absolutely vital to the success of any software project. 
However, scoping and specifying a project properly is far from simple.

Digital transformation, or taking manual or older client-server-based systems and improving them through incorporating AI and machine learning into these systems, can be hugely beneficial. However, to succeed, these projects have to be specified properly, so all the requirements and goals of the project are incorporated into the design process. 

McGuire points out that the software industry is littered with failed projects, many of which have resulted in losses running to hundreds of millions of pounds. 

The Herald:

John McGuire, founder and managing director of AI and machine learning specialists, Pulsion Technology


In fact, studies show that around 70% of digital transformation projects fail. The vast majority of these failures occur because the project was not properly specified in the first place. As a result, the delivered projects failed to achieve the hoped-for outcomes, McGuire notes.

In order to improve the process of specifying the requirements of a new software project, particularly one involving AI and/or machine learning, McGuire and his colleagues have built a design tool called Requiment 

“What we addressed with Requiment was the way misleading assumptions lurk in the gaps that are created when projects are not properly specified. People assume that the system will do a particular thing or address relationships between the data in a certain way, but they fail to articulate this properly.

"What our tool does is to use AI to guide the user through the requirements-gathering process. This results in the creation of a detailed and coherent requirements document and dramatically reduces the likelihood of a project being misspecified,” he explains.  

AI-driven requirements gathering and management. Save time and money by capturing your requirements right first time. Start a 30-day free trial today. Visit www.requiment.com


Incorporating AI into a tool that mimics the job of a business analyst helps to ensure that each stage of the specification process is logical and coherent. 

“Humans miss things, even skilled humans. By making the whole process AI-driven, we ensure that organisations work through the process of requirement gathering much more efficiently. 

“This is not just about the system asking you what you want and then documenting it. Our Requiment tool asks the questions that a top business analyst would ask and rapidly builds the screens that reflect what the final system, based on that information, will look like,” he comments.

This means that at a very early stage, the user community associated with the new software project, or the digital transformation exercise, can fine-tune the process or identify any false assumptions. 

As McGuire puts it: “The system helps you to see what you have missed in your initial requirements specification. Experience in our industry shows that if making an early change would cost you, say, £10, you can count on the fact that making that same change after the project has gone some way down the line, will cost you one or two orders of magnitude more.”

In other words, the big savings of implementing a tool like Requiment lies in avoiding the cost overruns that software projects so often incur. 

“The statistics show that in the software industry, projects with any degree of complexity associated with them, typically incur a 30% cost overrun. No one wants to see a project that was initially costed at £100,000 ending up costing £130,000,” he says.

More importantly, the IT sector is chronically short-staffed. Skilled specialists are hard to come by. So ensuring that a project that is initially scoped as requiring a team of three does not end up needing a team of four, can be vital if major delays are to be avoided.