Confronting the data centre talent crisis

Recent figures from Uptime Institute suggest that almost two thirds (63%) of data centre operators are struggling to attract and retain qualified staff. 

This squeeze on engineering talent comes at a time of unprecedented demand. The data centre market is projected to have a 11.06% compound annual growth rate (CAGR) from 2026 to 2035. This is the result of a boom in generative AI innovation and adoption, which has put considerable pressure on existing compute infrastructure. 

Without enough qualified professionals to support an ambitious infrastructure build-out, service continuity and even safety could be at risk.  

Upskilling across the data centre lifecycle  

So, what kind of work needs to be done? From construction to maintenance, engineering talent has a central role to play in the present AI revolution. The skills required to deploy new data centre infrastructure depends on what phase of a project a worker is involved in.  

Planning and design require high levels of technical expertise. There is little margin for error: you need people who can produce accurate single line diagrams that meet all the necessary design specifications. Professionals also need to understand the physics of data centres, from air flow to heat management. Often this requires Bachelor’s and Master’s degrees, which take years to complete and considerable investment.  

Building physical infrastructure comes down to actioning plans in line with particular standards. This is still a critical function, and installation entails following a particular manual to the letter. However, the threshold for training is generally lower. It demands specific certifications that can be acquired after high school or equivalent education. For example, some electrical certifications take just 2 or 3 months to complete. 

Optimising and maintaining data centres is the least skill-intensive work of these phases. It’s still skilled labor. But data centres have relatively few moving parts and single points of failure. As long as they are designed well in the first place and appropriate procedures are followed, many systems won’t fail for 20 years or more. As a result, upskilling in this area is relatively quick.  

Nurturing the next generation 

Data centres demand talent across the engineering spectrum, from designers and builders to maintenance engineers and electricians. With this skills matrix in mind, organisations must assess what the best approach to training and retaining talent is for them. This varies depending on their size and appetite for managing operations in house, as opposed to working with third-party partners and institutions. 

Hyperscalers are the ones leading the charge when it comes to data centre engineering talent. They have the resources and the global footprint required to nurture talent. This includes everything from sponsoring people to undertake 3 or 4 year programs before hiring them, to upskilling individuals who show high engineering potential in their maintenance work.  

Apprenticeships and in-house training have a key role to play in enabling firms to progressively support talent from high school onwards. They give people hands-on experience straight away. From there, they can develop into new areas based on their interests and aptitudes. 

Tier-one colocators are also driving change. They often offer internships and investment programs that similarly nurture talent and prepare them for the challenges of data centre operations. 

Tier-two and three colocations are generally less programmatic in their approach, as they have less capital to invest in these areas. Instead, they primarily hire talent from competitors. That means they tend to pay a premium during the hiring process, but benefit from the training and certifications employees have acquired up to that point. 

Despite the work being done, there’s still a shortfall – one that cannot be filled by traditional STEM programs or internships. Industry and government must do more to encourage talent into suitable pathways.  

One potential growth area is trade schools that focus on the specific skills necessary to work with data centres. Not everyone can study for multiple years, so supporting, partnering with, and actively hiring from these institutions will help make the most of the talent available. 

Future-ready fixes 

Fortunately, the same technologies that are placing heavy demands on data centre infrastructure can help address the talent shortfall. Automation and AI-driven insights can relieve data centre staff of many routine and manual processes, enabling them to focus on the tasks that matter the most. AI tools also make it possible to monitor workload distribution, dynamically adjusting resources to meet real-time demand, and identifying where efficiencies could be made.  

As well as managing power, AI is a great integrator. It can aggregate and analyse data across disparate systems to help employees make the right trade-offs and identify opportunities for cost saving across the business. Previously time-consuming decision-making processes, like Scope 1, 2 and 3 emissions reduction, can be augmented and even automated with the help of AI. 

New innovations can also help operators to upskill existing talent. Digital twins can be used for training, just as pilots use flight simulators to learn to fly, engineers can practice assembling, installing or repairing data centre infrastructure in virtual environments. This doesn’t replace the need for formal training and accreditation, but it can support development with interactive scenarios in controlled virtual environments. 

In parallel, new operating models are reducing the burden on onsite engineers thanks to modular, prefabricated, and containerised designs. Operators give their specifications to a manufacturer, and the build takes place in factories that benefit from automation and advanced robotics.  

From there, mobile substations are transported to sites on 18-wheelers – all that’s left to do onsite is lock the containers to concrete slabs and connect them to a power source. It’s a “plug and play” solution that dramatically reduces the inefficiency of conducting complex work in the field. 

A joined-up approach 

Data centre operations require a complex matrix of interrelated skills across both IT and OT. New digital technologies are being deployed alongside increasingly sophisticated cooling and energy management systems.  

If implemented effectively, this offers considerable efficiency gains, helping operators reduce costs, hit sustainability targets and meet compute demand. It could even bring us closer to autonomous operations that require reduced human input.  

The talent crisis isn’t going anywhere, and we need to use every tool at our disposal to tackle it. That means investment across training, apprenticeships and higher education, along with labor-saving innovations. This is not a case of replacing jobs. It’s about finding ways of getting the job done in the face of an acute shortage.