Authored by Tom Perkins, Vice President, Data Center Operations

The conversation around artificial intelligence tends to focus on algorithms, compute power, and cloud architectures. But behind every AI workload is a physical facility, and behind every facility is a team of skilled professionals who design, build, and run it. As AI infrastructure scales at unprecedented speed, the data center industry is confronting a workforce challenge that may be more pressing than any power or land constraint: a deepening skills gap.

A Workforce Built for a Different Era

The traditional data center workforce was built around IT technicians, network engineers, and facilities managers with expertise in conventional server environments. That foundation remains important, but AI is rewriting the job description at every level. According to Deloitte, 63% of data center executives cited a shortage of skilled labor as their single greatest obstacle to talent acquisition. Between 2023 and 2025, data center job postings surged 64% – far outpacing the 4% growth seen across the broader economy – while postings for electrical technicians alone climbed more than 180%, Deloitte also reported.

The Physical Side of AI Gets Overlooked

There is a tendency to think of AI talent shortages in terms of data scientists and machine-learning engineers. That’s only part of the picture. Randstad‘s analysis of over 50 million global job postings found that since 2022, demand for HVAC engineers has increased 67%, robotics technicians are up 107%, and electricians are up 18%. Hiring a skilled tradesperson now takes an average of 56 days, and the United States will need approximately 300,000 new electricians over the next decade on top of replacing the 200,000 expected to retire, according to FORTUNE. Uptime Institute found that 51% of data center operators struggled to find qualified candidates in 2024, with the largest shortfalls in junior and mid-level operations roles.

New Skills for New Infrastructure

As data centers evolve to support AI, the skills required to operate them are changing just as fast. Legacy cooling systems are giving way to liquid and immersion cooling. Power densities that once averaged 8 kW per rack have more than doubled to 17 kW, with some AI training environments pushing beyond 80 kW. Managing this infrastructure demands a new blend of expertise: part mechanical engineer, part IT architect, part energy specialist.

The World Economic Forum notes that companies prioritizing upskilling and reskilling their existing workforce, providing clear pathways for employees to grow and transition into more advanced roles, are better positioned to retain institutional knowledge and build a more adaptable team. This approach is especially critical given that nearly 60% of in-demand data center skills are actually non-technical, spanning problem-solving, critical thinking, and adaptability – qualities that can be developed in current employees rather than sourced exclusively from outside. Organizations that invest in competency-based training and cross-functional career pathways will be far better positioned than those chasing talent in an increasingly competitive market.

What This Means for Colocation Customers

This talent landscape has direct implications for enterprises evaluating their infrastructure strategies. Operating an in-house data center is not just a capital challenge – it is a workforce challenge. Recruiting and retaining the specialists needed to manage modern AI-capable facilities are costly and time-consuming, and when those roles go unfilled or turn over, operational continuity is at risk. Colocation providers absorb that burden. At 365 Data Centers, our team of operations professionals stays current with the evolving demands of high-density, AI-ready infrastructure, and our 24/7/365 staffing model means knowledgeable support is always available, giving customers access to deep operational expertise without building an in-house team in one of the tightest labor markets in recent memory.

Looking Ahead

The skills gap in data center operations is not going away on its own. It will require investment from operators, training partnerships with academic institutions and trade programs, and smarter workforce development strategies across the industry. But for enterprises looking to run AI workloads today, choosing the right infrastructure partner, one with experienced teams and the operational depth to manage what is coming next, is one of the most practical steps available. The AI era demands more from data center infrastructure. It demands equally more from the people who run it. Understanding that connection is where smart infrastructure decisions start.

Learn more about how 365 Data Centers can support your infrastructure needs at 365datacenters.com/contact.