Data centers face shortage of professionals with experience

The following article is a collaboration between IT Brew and commercial real estate news platform Bisnow.

It was Thanksgiving week in 2022—a time when most employees are hard to find, let alone critical pieces of data center power infrastructure. And Kurt Bogle needed a transformer.

Bogle, then a senior regional director of critical facilities operations at NTT Global Data Centers, oversaw seven operational buildings around Chicago and Virginia and a staff of about 150 people. After a transformer failed a test, Bogle had to find a replacement and coordinate a swap with the onsite managers—all without disrupting service.

Thankfully, Bogle found a spare, but his experience illustrates the importance of systems thinkers at data centers who know how to acquire the resources necessary to maintain data centers’ high uptime expectations. Given supply-chain limitations and a steady buildout of data centers, such expertise is more crucial than ever—and there are signs a labor shortage is underway, threatening Big Tech’s AI ambitions and the reliable operation of mission-critical facilities.

“It’s going to get stretched,” Bogle said. “There is a need to continue to develop, build, mentor these leaders so that they can take on the roles of today and tomorrow.”

Already, more than half of data center operators now report difficulties finding qualified candidates for vacant roles, a significant jump from last year, according to a report published recently by industry think-tank Uptime Institute. Firms are competing for a limited pool of electricians, mechanical technicians, and other highly specialized operations staff needed to keep increasingly complex facilities up and running around the clock.

Big Tech’s AI ambitions are accelerating the race to build new capacity, creating a labor shortage not only for the construction workers building the facilities, but for the qualified workers needed to staff data centers once they are operational.

Addressing these workforce woes for operational data centers isn’t as simple as just hiring more people. Although firms across the industry are increasingly confident in their ability to train and hire junior-level workers across a range of roles, the labor shortage is far more intractable—and increasingly acute—when it comes to workers like Bogle: experienced, senior-level industry veterans who are essential for reliable data center operations.

Without an adequate pipeline of these seasoned data center pros, experts say, systems in finance, healthcare, and other sectors that rely on these mission-critical facilities face a higher risk of outages and a cascading “blast radius” of failures that could cause widespread disruption and billions in economic damage.

Even as companies scramble to develop career pathways and training programs to cultivate a new generation of senior-level leaders, the process could take years to bear fruit. And that may be time that the industry doesn’t have.

“There is an expert shortage because we are not replenishing the talent pipeline,” said Rose Weinschenk, a researcher at Uptime Institute who coauthored the group’s report. “[That] is exactly why we see these companies trying to keep pushing retirement for a lot of these experts that they already do have further and further away and just clinging to them as long as they can.”

Rapid growth. The US already has more than 4,000 operational data centers, with nearly 3,000 additional facilities either under construction or planned, according to Axios. At the same time, these facilities are getting dramatically larger, with campus-scale projects increasingly measured in gigawatts rather than megawatts.

Each of these facilities requires teams of operators, facilities engineers, electricians, and mechanical technicians. As a result, demand has surged for workers with these speciality skillsets, placing further strain on a labor pool that was already stretched thin before the AI boom.

Compounding the industry’s broader labor pinch is the sudden geographic expansion of data center development.

Until the last few years, nearly all facilities were concentrated in a handful of established hubs, the largest being Northern Virginia. Large-scale data centers were rarely built outside of major metro areas. But as power has become scarce in traditional markets—and because AI training workloads don’t have the latency concerns that require other data centers to be located near large population centers—developers have increasingly pursued sites where large blocks of electricity are available.

This shift has fueled the rise of gigawatt-scale campuses for tech giants like Amazon, Oracle, and Meta in rural locations, with projects built or planned in places like Richland Parish, Louisiana; Abilene, Texas; Ellendale, North Dakota, and Lea County, New Mexico. Commercial real estate company JLL reported last month that 77% of data center capacity under construction is in “frontier markets,” not traditional hubs.

Established data center markets have existing experienced labor pools and talent pipelines, even if they are falling short of the sector’s current needs. But for rural projects, those ecosystems don’t exist at all, forcing operators to try to build specialized workforces from the ground up fast enough to deliver the speed to market demanded by tech giants amid the AI arms race.

“Our clients are hyperscalers or colocation providers, and they’re doing work everywhere, so when they ask us to do work in South Dakota, we don’t have people there,” said Kurt Haglund, chief operating officer of Virginia-based data center operations specialist Compu Dynamics. “Filling the work wherever it happens to be is probably the biggest challenge.”

Greatest need. Whether in Northern Virginia or North Dakota, the industry’s greatest hiring need is entry-level and mid-level operations roles: everything from security and equipment monitoring to electricians and facilities engineers.

Operators responding to Uptime Institute’s survey reported the greatest skill gap in electrical roles—which increased 13 percentage points since 2023—plus junior-level operations, operations and mechanical positions.

Industry leaders acknowledge they’re playing catch-up when it comes to building talent pipelines for these roles. Many data center firms are now investing heavily in their own training programs and workforce partnerships colleges.

Compu Dynamics is among the companies that has shifted its approach. Rather than searching only for workers with existing data center experience, the firm has devoted significant resources over the past 24 months to recruiting candidates with transferable technical skills and providing the specialized training needed to work in mission-critical data centers. It’s a strategy Haglund says has proven successful.

For Applied Digital, developing a hyperscale campus in Ellendale, North Dakota, required building a workforce from scratch in a region with no existing data center talent. According to CEO Wes Cummins, the company ultimately hired up to 85% of its employees locally, relying on in-house training and training partnerships with equipment vendors and with nearby colleges to prepare workers for highly specialized roles.

But while data center operators might be succeeding at building talent pipelines for junior positions, they face a far more difficult challenge in addressing the acute shortage of industry veterans qualified for senior-level roles.

The need for senior leadership is critical across every aspect of data center operations. In a mission-critical environment with little room for failure, seemingly mundane tasks require experienced pros who understand how to minimize risk and avoid catastrophe.

Even transporting equipment requires a veteran’s understanding, suggested John Ahlering, AI engineering and team lead at data-center services provider Salute. With close to three decades in data centers, Ahlering creates task hazard assessments when guiding his team through the choreography of moving heavy, expensive equipment—or even a simple cable.

“One of the most dangerous parts in a data center where damage can happen is doing transportation. I’ve learned that in 30 years,” Ahlering said.

The veteran data center workforce isn’t just being spread thin; its ranks may actually be shrinking due to the “Silver Tsunami,” with a disproportionately large share of workers approaching retirement age. As these veteran operators, engineers, and managers leave the industry, companies risk losing critical institutional knowledge exactly when it’s needed most.

This kind of operational expertise can’t be developed overnight.

Applied Digital’s Cummins says senior positions account for the roughly 15% of his workforce he can’t hire locally at his Ellendale campus. Instead, he is forced to convince senior leaders from other firms to relocate to North Dakota.

“There’s no training for experience,” Cummins said. “You need enough people that have had the experience, that have had an issue at a data center, that know how to deal with it, that know how to fix it.”

Other firms have had to lure employees with dramatic salary increases, with some roles now commanding up to $300,0000 annually. Last year, 28% of data center operators reported having staff hired away by competitors, according to Uptime.

As a result, operators across the industry are increasing salaries to retain the employees they already have or to prevent aging workers from retiring. It’s a bidding war that experts say may address operators’ short- term needs, but does little to solve the industry’s increasingly acute labor woes.

“I have been solicited many times,” Ahlering admitted. “They need those [kinds] of guys who can take the daily ops and make it work, and keep the systems up and running; see the problem before it happens, mitigate this, and take ownership. That’s the key of my job. I have to own everything.”

Failure to develop more experienced workers like Ahlering could result in diminished data center reliability, industry leaders say, which carries significant consequences extending well beyond the data center and tech sectors. Data centers support a range of essential services, including financial, government, and medical environments. Almost 20% of data center outages result in at least a million dollars in economic damage, according to Uptime.

In other cases, keeping data centers up and running is a matter of life and death. For example, Ahlering works in a high-security facility that provides critical services to tenants that include medical software systems. “If we were to have a system go down where the doctors do not know the patient’s allergies, they can’t prescribe medicine. These are very critical services we are responsible for,” Ahlering said.

Growing the future. The industry’s shortage of experienced workers is, in many ways, a consequence of its failure to retain and develop the employees it already hired. Even as operators brought new workers into the sector, many neglected to create clear career pathways that would keep those employees in the industry long enough to become the senior data center pros now in critically short supply.

According to Compu Dynamics’ Haglund, turnover rates of 30% have been common across the industry, reflecting a prioritization of replacing junior workers while neglecting retention efforts needed to build a roster of experienced veterans. Only recently have operators begun investing seriously in formal career development and employee retention initiatives.

At Compu Dynamics, this has meant replacing a traditional HR department with a dedicated talent management division focused on onboarding, mentoring, professional development, and employee culture. New hires are paired with mentors, organized into smaller teams, and given access to structured learning programs that show clearly defined career pathways and how employees can advance within the company.

This strategy has paid off, according to Haglund. Since fundamentally overhauling its approach to employee development, the company has reduced turnover from around 30% to around 10%.

“Our goal is if we can’t find raw talent with a lot of experience, then we find raw talent that we can nurture and grow into the people that we’re going to need in the future,” Haglund said.

Areas with an established data center presence have also begun standing up skills-training initiatives. Virginia, home to almost 700 data centers, has state-supported programs like FastForward. This initiative, established in 2016, offers ways to achieve credentials in skills ranging from commercial driving to welding—many essential to data center buildouts and maintenance.

While those skills are not senior-level ones, Randall Stamper, associate vice chancellor for career education and workforce programs with the Virginia Community College System, suggested it’s a start: “We can get people what they need to advance into, let’s say, middle management.”

Amazon Web Services, which has invested over $119 billion in Virginia data center buildouts and maintenance since 2011, including capital and operational expenses related to AWS’s data center infrastructure, has used many of FastForward’s participants to build its data centers, and needs more to run the facilities long-term.

“As you grow out entry-level technicians, now you have to grow a leadership, essentially hierarchy within that organization, because one of the challenges here is we have a huge influx of new people coming into the industry, and so the people who are already in it need skills that are more management skills,” Nicholas Lee-Romagnolo, principal at AWS Economic and Workforce Development, told us. “We need people who have the experience, and that can be a real opportunity for people who are in the data centers.”

These efforts may ultimately prove effective, but growing the ranks of senior leadership is a process that, by its nature, will take years. In the meantime, data centers are desperate for those workers today.

Kris Beevers, co-founder and CEO of infrastructure management company NetBox Labs, sees pros who had been working in IT infrastructure moving into senior data center positions, like site-reliability engineers migrating to data center work, as a “trial by fire.” Similarly, firms like Compu Dynamics are hiring workers from other fields with “transferable” skill sets who, with some training, can effectively fill certain roles traditionally filled by more experienced workers.

Automate this! Beevers also sees automation playing a significant role by helping to reduce workforce needs; for example, assisting with tasks previously conducted by senior workers, such as determining a data center’s bill of materials and assigning and running workloads once the components are plugged in.

“You really shouldn’t have to have an engineer at a drafting table figuring out, ‘How am I going to lay out these racks so that the floor doesn’t collapse?’ That’s an entirely automatable problem,” he said. “It is impossible to build the kind of data center scale that we’re hearing about in 2026 or even the last few years without substantial automation.”

Additionally, some within the industry say that the rapid evolution of data center systems is narrowing the knowledge gap between senior employees and their less experienced peers.

AI has dramatically changed the IT hardware within data centers, increased rack densities and forced the widespread adoption of liquid cooling and other technologies operations staff rarely encountered just three years ago. As a result, even long-time workers need to update their own skillsets significantly. Ahlering, for example, has upskilled on liquid cooling and its ability to maintain rack temperatures.

But industry leaders say experienced workers bring value beyond just their technical knowledge—and beyond what automation can provide. Without a kind of “guiding light” from senior leaders who have seen a challenge or two, “you get more of a lag in decision-making because you have people that are struggling to make those decisions,” Bogle said.

While automation may help address short term workforce needs, Uptime’s Weinschenk said it may ultimately hinder the skill development of junior employees who could become leaders in the future.

“What tends to happen is then there won’t necessarily be a holistic understanding of the systems,” she said. “The skill development seems to be limited to a small number of people who are deemed to be future subject matter experts, and then you have a lot of turnover for these junior-level roles who are doing things like walking around, checking to make scanning QR codes, and making sure that everything is where it’s supposed to be.”

Bogle has prioritized this kind of skill development with his more junior colleagues. He likes a crawl-walk-run approach with his staff: have someone run a project; then lead a team; then lead multiple teams. That kind of 1-on-1 leadership, built on years of problem solving, imparts unique knowledge that can prove harder to find than a spare transformer on a holiday weekend.

“What’s lost if you don’t have the high senior leaders is that mentorship and experience that comes from that mentor,” Bogle said.