Michael Beagan, MD of TES Power, explains why solving the industry’s growing skills shortage will be just as important as securing the power and infrastructure needed to support AI.

The AI boom has pushed the data centre industry into a generational growth spurt. AI companies are clamouring for the necessary compute capacity to train and run their generative AI models as user numbers pass the billion mark. It’s easy to get caught up in the sheer scale of it all – the multiple gigawatts of planned construction, the recent 1,800% spike in Samsung’s profits from chip sales, the projected $4.8 trillion value of the AI market by 2033.

Building the infrastructure necessary to support the AI revolution is widely acknowledged as a monumental challenge of both logistics and design. Yet behind the headlines surrounding power availability, compute capacity and hyperscale investment lies another challenge that may prove equally important: people.

Somewhat ironically, in this new golden age of automation, it’s all about people.

The data centre skills shortage is increasing demand for skilled workers at a time when the industry is already contending with the challenges of scaling along unprecedented growth trajectories and overcoming new design challenges posed by AI-first digital infrastructure. As AI-driven demand accelerates, success will depend not only on capital investment and engineering innovation, but also on the industry’s ability to attract, develop and retain the talent required to deliver increasingly complex projects at unprecedented speed.

Even in the age of AI, rising to the technical challenges of scaling, shorter development timelines, powering new kinds of AI workloads, and every other thorny problem with which the industry is grappling eventually comes back to people.

The skills shortage

As demand for AI data centres soars, organisations throughout the supply chain are running into an industry-wide skills shortage. According to the Uptime Institute, at least 250 global data centre projects exceeding 100MW in energy demand were announced between 2021 and 2024. Finding people with the skills and experience to design and deliver these highly complex technical environments has quickly become a serious challenge.

Even a few years ago, the Uptime Institute identified that operators were having difficulty sourcing candidates as demand outstripped supply in the early days of the AI boom. Most companies surveyed (58%) had difficulty finding qualified candidates for open jobs, and more than half (55%) reported difficulty retaining existing staff, with 40% of those citing staff being hired away by competitors in the sector as the primary cause. Engineers, electricians, plumbers, commissioning agents and construction workers all command significantly higher salaries today than five years ago.

This shortage is not a temporary blip that will resolve itself once a few new training courses come online. It is a structural problem, and it calls for a structural response.

Apprenticeships and building the next generation

Apprenticeships have re-emerged as one of the clearest routes to closing the skills gap. Rather than competing for an already stretched pool of experienced engineers, a growing number of data centre businesses are building their own talent from the ground up.

Effective apprenticeship models do more than teach a single trade. Data centre delivery spans mechanical and electrical disciplines, manufacturing, testing, commissioning and project management, so apprentices who rotate through several of these functions early in their career can develop a far broader understanding of how a finished system comes together. That breadth matters more than it once did. AI-ready facilities are more complex than the data centres of a decade ago, and the people who build them benefit from seeing the whole picture rather than a single stage of it.

There is also a potential retention benefit. Apprentices who progress through structured, varied pathways and can see a credible route into engineering or leadership roles may be more likely to stay with the employer who trained them. Employers who treat apprenticeships as a long-term investment rather than a recruitment shortcut are also more likely to develop a sustainable pipeline of skills.

Transferable skills and knowledge retention

A less visible but equally pressing problem is the aging of the existing workforce. Much of the specialist knowledge behind data centre electrical and mechanical systems still sits with individuals rather than with documented processes. When that experience is concentrated in a small number of long-serving specialists, a single retirement or resignation can take years of institutional knowledge with it.

Addressing this means treating knowledge retention as a deliberate discipline rather than something that happens informally on the factory floor or the job site. Centralising technical knowledge and building systems that make expertise accessible across teams, rather than locked within individual specialists, gives every project the benefit of the organisation’s full experience, rather than just that of whoever happens to be assigned to it. It also gives newer employees a faster, more structured route to competence, rather than years of trial and error.

Transferable skills matter here too. Electrical and mechanical expertise developed in one part of the business, whether in switchgear assembly or panel wiring, can often be applied elsewhere with the right training. Employers who invest in cross-training reduce their dependence on any single skill set while giving employees opportunities to develop broader and more resilient careers.

Breaking down siloes through cross-functional development

Manufacturing and electrical engineering are highly specialist disciplines, and in any organisation racing to deliver complex systems at speed, there is a real risk that these functions become disconnected from one another. Silos slow decision-making and make it harder for employees to see how their work contributes to the finished product.

Breaking down these silos brings a twofold benefit. Information and materials can move more smoothly through the business, while employees gain greater visibility of how their individual contribution fits into a larger, collective effort. That sense of ownership is difficult to create through policy alone, but it is more likely to develop in organisations that deliberately encourage visibility across functions.

Practical changes employers can make

None of this requires reinventing how a business operates overnight, but it does require employers to treat workforce development as a strategic capability rather than simply an HR initiative. Several practical changes are worth considering.

Giving employees a genuine voice in how the business operates is one of them. Some manufacturers have trialled changes such as a four-day working week for shop-floor employees in response to staff feedback. Flatter management structures that keep leadership visible and accessible on the factory floor, rather than removed from day-to-day operations, can support the same goal: making it easier for problems and ideas to surface quickly, rather than getting lost in layers of hierarchy.

Physical spaces matter more than they are usually given credit for. Shared spaces designed to encourage conversation, whether that is a canteen, a break area or an open-plan design office, can shape how readily people share knowledge and solve problems together. It sounds like a small detail, but the way employees use and adapt these spaces can offer a useful indication of the culture underneath.

The human side of the AI boom

The AI era is placing unprecedented demands on the data centre ecosystem. Investment and technology will be critical factors in determining how quickly the industry can scale, but the ability to meet future demand will ultimately depend on the people responsible for designing and delivering that infrastructure.

Apprenticeships, transferable skills, deliberate knowledge retention and cross-functional development are not simply nice-to-have additions to a workforce strategy. They are becoming increasingly important for organisations attempting to deliver at the pace the AI boom demands, particularly as the industry competes for a limited pool of experienced talent.

The next phase of AI infrastructure growth will undoubtedly be driven by technology and power. One of the sector’s most important competitive advantages, though, will remain deeply human: the ability to attract and retain the people capable of turning ambitious plans into operational reality.