As today’s data centers powering AI infrastructure become more complex, the tech industry must provide ways for people to get the necessary skills. AI data centers require a large amount of compute power to train advanced AI models, and a shortage of workers makes filling data center roles more challenging.
For example, in Deloitte’s 2025 AI Infrastructure Survey, 63% of data center executives reported a shortage in labor related to data centers as the top obstacle in finding talent. The survey authors discussed the challenges in finding workers in both the power and data center sectors, which are key to AI infrastructure growth.
“The challenges are multifold: Both sectors are scaling at the same time and rely on the same core workforce, many roles increasingly require digital and AI skills, and training can be lengthy,” the Deloitte authors wrote.
William Porter, data center operations manager at AWS, notes a shortage of key roles in data centers that include electricians, fiber splicers, linespeople and engineering operations technicians.
“While much attention is paid to the ‘tech talent gap’ for software developers, there is also great need across the economy to support the skilled technical workforce — people performing hands-on jobs that require significant technical knowledge,” Porter says.
Deloitte expects data center power demands to climb from 47 gigawatts in 2025 to more than 176 gigawatts by 2035, and that means data centers will need skilled workers to power AI infrastructure.
In addition, data centers are growing with 38% of Americans now residing within 5 miles of a data center, according to the Pew Research Center.
“The rate at which they are growing is massive, and if you look at the design and build phase — all the way from installing the [equipment], electrical installations, plumbing, construction — these are the areas where massive explosion of talent or workforce is required,” said Vineet Thuvara, chief product officer at Fluke, a 78-year-old test and measurement company. electrical testing and measurement tools.
For data centers, managing power generation is the biggest challenge, according to Thuvara. Compute spikes on short notice can drive up power consumption.
Electrical, Chemical Engineering Skills in Demand for AI Data Centers
For professionals pursuing an electrical engineering path in data centers, Thuvara recommends learning the basics of electricity, power and battery management.
These tech pros may focus on construction and cabling and HVAC systems, he says.
Tech professionals should pursue training in mechanical engineering and electrical engineering, but for data centers they don’t need a full engineering degree, according to Thuvara.
“They could also be technical college students who are going through a skills-based program instead of an engineering degree-based program,” he says.
In addition, tech professionals preparing to work in AI data centers should learn some chemical engineering and how environmental monitoring works in a data center, including air quality.
“There are different kinds of scientific aspects that are at play depending on the area you try to focus on,” Thuvara says. “For example, liquid cooling in terms of heating that happens in the data center is a different kind of thermodynamics that is at play.”
He recommends learning the basics of physics and chemistry, but tech pros do not need to be chemical engineers, according to Thuvara.
“You need to understand [how] to make sure you're making the optimal decisions for the data center,” Thuvara says. “Battery management is another massive chemistry piece.”
Here are some ways that three organizations are training tech professionals with data center skills.
Johnson Controls Launches Data Center Training Institute
In May, thermal management company Johnson Controls announced it would launch a data center training institute at its Advanced Development and Engineering Center (JADEC) in New Freedom, Pennsylvania. The facility is slated to open in early 2027 and offer hands-on training to service technicians in tackling the demands of high-density data centers.
“Data centers don’t run without skilled technicians, and as technology advances, demand for that expertise only continues to rise,” Buddy Saucier, vice president of HVAC service operations at Johnson Controls, said in a statement. “This training center will set technicians up for long, meaningful careers while delivering the reliability and support customers depend on every single day.”
The JADEC facility extends over 7 acres and specializes in thermal management technologies. In addition, Johnson Controls partners with Lincoln Tech across its 18 campuses to prepare tech professionals with essential AI data center skills. Lincoln Tech provides IT training in several states.
Fluke Trains Future Electrical Engineers
Network cabling is, of course, an important infrastructural piece of an AI data center, and network testers are in demand, Thuvara says.
Companies such as Fluke also need power quality analyzers and battery testers in their AI data centers.
Thuvara works with junior apprenticeship training centers (JATCs), for which two electrical training bodies develop the curriculum: the International Brotherhood of Electrical Workers and the Electrical Testers Institute (ETI).
Meanwhile, Fluke is partnering with Washington State University to set up labs, build curriculum and hire early-in-career technicians.
“We are working with their faculties to make sure our tools are available, training them up front,” Thuvara says.
The training will be valuable for tech professionals because electrical measurement tools are important for AI data centers.
“These devices make sure that the devices they are using to measure the infrastructure are generating the right amount of output,” Thuvara says. “So it gives you accurate data on the output of the infrastructure.”
AWS Upskills on Data Center Infrastructure
Like Fluke, AWS is also working with the community to train tech professionals on the data center skills they need.
“There's a whole community of local businesses that build and maintain this infrastructure, and our training programs are designed to feed that entire pipeline,” says Chris Affrunti, director of workforce development at AWS. “That's why we're partnering with local schools and community colleges, launching paid pre-apprenticeships, and creating industry credentials that open doors whether someone ends up at AWS or with one of the many local contractors who power this work."
AWS offers upskilling programs and collaborates with local workforce development and education organizations to train technologists for AI data centers. They consist of a technical workforce in the construction or skilled trades, fiber optics, data center operations and energy generation and distribution.
In addition, along with telecommunications equipment supplier Sumitomo Electric Lightwave, AWS offers courses such as the AWS Fiber Optic Fusion Splicing Certification Course, which provides hands-on skills in fiber-optic installation and repair. These skills are key components for building AI data centers.
AWS also offers one-day events such as the Information Infrastructure Workshop (I2W) for educators and workforce leaders to prepare for data-center related topics such as cloud computing and the information economy.
The company also works with colleges to coordinate academic programs with data center roles. For example, AWS has partnered with Reynolds Community College in Richmond, Virginia, to allow students to complete a two-day fiber-optic installer (FOI) course. Fiber-optic cabling provides the power for AI in data centers.
AI-focused data centers require about 36 times more fiber than “traditional” CPU server racks, Rahul Puri, CEO of the Optical Networking Business at STL told Fierce Network last December. That’s why efforts by hyperscalers like AWS to train tech professionals will be valuable.
