From Chaos to Control: Tackling AI’s Toughest Obstacles with a Process-First Mindset

Throughout every major epochal change, companies have always needed to adopt a process-oriented mindset. The same is true today with the adoption of Artificial Intelligence. AI is not merely an extra technology or a passing trend, but rather a way to rethink organizational structures, workflows, and internal culture. According to most industry experts, such as Nvidia’s CEO and the ever-quotable Elon Musk, we will see a radical transformation of our society by the end of 2025, impacting both goods and services industries, although in my view, services are already feeling a significant impact. Drawing from my own consulting experience in different sectors, comparing industries like Automotive and manufacturing, which are often reluctant to revise their workflows, with areas like Luxury and Fashion, which have taken more of a top-down roadmap approach, it seems that there will be two broad ways of handling this transformation. Those who swiftly and effectively incorporate processes and involve people from the outset will come out ahead.

At a time when AI in financial markets is experiencing what could be perceived as a “bubble” similar to the .com era, it’s worth remembering a lesson from history: every bubble has, in its own way, sparked benefits for society that have left an indelible mark on our industrial and social landscape.

For years, especially in my past experience as an IT consultant, the traditional manufacturing industry struggled the most with the topic of digitization, introducing sporadic software systems or automation solutions without fully integrating them into a broader strategic plan. The result, inevitably, was slowdowns and obstacles: employees struggled to accept new technologies, IT teams had to battle siloed data, and governance became muddled by unclear responsibilities. One obvious outcome is that projects often fail due to a lack of buy-in from key users (those deeply involved in the processes and supportive of change). Without that shared commitment, tools may go unadopted, user acceptance testing (UAT) may be rejected due to insufficient knowledge of the solution, or worse, because of flawed logic built into the implementation.

AI, in contrast, highlights the need for well-structured processes precisely because, more so than other innovations, it relies critically on a company’s informational assets. If data isn’t consolidated, cleaned, and available in a consistent manner, AI risks producing misleading results, triggering internal doubts and criticism. Beyond the importance of reliable data, the very nature of AI demands a culture of collaboration, continuous feedback, and progressive improvement. Never before has the need been greater for a structured, company-wide knowledge base, which, within the TOGAF framework, is referred to as the Architecture Repository. Indeed, the vast majority of AI products can be trained using company documents such as manuals, schematics, and other materials. Having a well-organized corporate knowledge base is the foundation for successfully adopting such technologies.

This is where the importance of a top-down approach becomes clear. When AI is adopted in a “bottom-up” fashion, initiated by a single department or a small group of enthusiasts, there is a heightened risk of strategic misalignment. The outcome is often “spot” solutions that work well in a limited scope but can’t be integrated across the enterprise.

Another key factor is change management, which becomes even more challenging in the context of AI. People in the company may fear stronger oversight or the replacement of certain roles, fueling resistance. It takes concrete effort to demonstrate tangible benefits, explain how the technology integrates into everyday operations, and provide adequate training pathways. Without a cultural shift, AI becomes a source of anxiety rather than an opportunity. That’s why it’s important to start at the top: leadership must define objectives and priorities, build upskilling or reskilling programs, and align every level of the organization around principles of transparency and collaboration. I also recommend checking out interviews with Nvidia’s CEO, who articulates a visionary perspective and boils it down into simple terms: if a company earns more thanks to AI, it will create different jobs. In other words, there’s no point in being afraid, unless, perhaps, you’ve chosen not to learn about AI at all.

Are you ready?