AI Must Transform Entire Businesses

Buying an artificial intelligence tool does not, by itself, constitute business transformation. “Artificial intelligence is effective only when it is integrated into practically the entire business process,” Krešimir Mlinarić, CEO of Span Evolve, tells ICTbusiness Media – ICTbusiness.info. For him, the central challenge is organisational: businesses must rethink how work moves from one stage to another.

Mlinarić distinguishes between employees feeling more productive and a company achieving a measurable improvement in its results. A person may complete an individual task faster without changing the bottlenecks, approvals or handovers that determine the performance of the wider process. “Personal productivity is not the same as corporate productivity,” Mlinarić explains.

That distinction is why he rejects an approach in which AI is treated as another software licence. The real opportunity lies in redesigning how tasks are carried out, how information moves and how decisions are made. The value of automation depends on whether those changes connect across departments rather than remain confined to isolated experiments. “You have to look at the entire process,” Mlinarić says.

He also considers the impact of AI broader than that of many previous technological changes. Because it can affect almost every business function, responsibility cannot simply be assigned to the IT department. Senior executives need to understand the possibilities, establish priorities and lead the organisational changes required for implementation. “This transformation must be driven by management,” Mlinarić stresses.

In that sense, AI is a management issue as much as a technology issue. Introducing tools without changing workflows may produce encouraging demonstrations but limited financial or operational benefits. Mlinarić argues that the companies making real progress will be those prepared to reconsider established practices rather than merely accelerate existing ones.

He also identifies AI agents and emerging transaction models as indications of where the technology is heading. These developments increase the potential for systems to carry out sequences of tasks rather than assist with a single step. “The change is much bigger than using a chatbot,” Mlinarić notes. The implication is clear: the organisations best placed to benefit will be those that integrate AI into their operating model, not simply those with the largest collection of tools.