Meet Two Twenty-Something Founders Helping Power the AI Goldrush

The AI “gold rush” hasn’t emerged in isolation. Rather, it has coincided with a growing need for complementary capabilities and infrastructure to support its development and adoption. This pattern isn’t entirely new. Similar dynamics have surfaced during previous waves of technological change. As each wave gains momentum, it often opens up adjacent business opportunities for those able to anticipate a need for additional support and services.

You don’t have to look far for examples. The computer revolution created a need for things like semiconductors and standardized hardware. The explosion of personal computing didn’t happen solo, either. It drove demand for affordable computer parts and more accessible software ecosystems. 

Entering the 21st century, the internet boom similarly created a need for data centers and routing infrastructures. It drove even greater demands for broadband access, server capacity and new security protocols.

You can find plenty of similar examples in the emergence of e-commerce, streaming, search and social platforms. When technology advances, it often gives rise to new layers of demand, along with opportunities for entrepreneurs who are prepared to address emerging needs.

Brent Whitehead and Matt Lohstroh of Giga Energy are two business founders who spotted an opportunity due to the growing need for American energy to power the artificial intelligence surge.

Building data centers from scratch in East Texas

Lohstroh and Whitehead were just 21 when they built their first modular data centers, welding and spray painting their early designs by hand in the oil fields of East Texas while students at Texas A&M. Bitcoin was on the rise, and rather than simply invest in cryptocurrency (like normal college students), they figured the secret to the biggest gains was starting from scratch.

The pair quickly learned what other operators have long understood: sourcing electrical infrastructure is a mess. They needed transformers, but lead times were outrageous. They needed switchboards, but customization was expensive. Every piece they needed for their burgeoning business was more complicated than the last.

“Everywhere we tried to grow, we ran into a broken supply chain,” said Lohstroh. “The equipment, the support. We knew there had to be other people out there like us, running into the same problems.”

So, as young people often do, they tried to do something impossible: manufacture their own industrial equipment from scratch. The duo called in reinforcements, Scott Pickering and Shiv Gupta, to round out their team. Gupta was tasked with rethinking the status quo of the supply chain. Then, they set out on their one mission: to learn how to make better equipment faster.

As the team scaled its operations, opening factories in Long Beach, CA and Houston, TX, Giga Energy began to attract new kinds of attention. Electrical distributors, tired of long waits and poor service from century-old legacy manufacturers, started calling. Giga Energy entered a new phase as a serious provider of electrical infrastructure, not only for its own data center projects but also for buyers in renewables, commercial and industrial and utility spaces.

Before long, the AI and hyperscaler companies, and their rapidly rising demand for energy, took notice.

Manufacturing better answers to new AI problems

AI giants require massive amounts of energy to function, and someone has to provide the infrastructure underneath that energy.

The reality of the situation is twofold. Financially, the United States has already made up its mind about AI investment. Without data centers, US GDP growth in the first half of 2025 was just 0.1%. 

Second, AI uses power on a large scale. By 2028, AI will consume as much power annually as a combined 22% of American households. As a result, AI is becoming an increasingly important factor in discussions about the future of both the U.S. economy and the American energy sector.

Enter another problem: the American grid as it exists today cannot handle this growth. One US government report from the summer of 2025 warned that within five years, blackouts across the country could increase by 100 times current rates without substantial energy improvements.

Those energy improvements will come from companies like Giga Energy. With its expanding American manufacturing presence and experience in developing in-house data center projects, Lohstroh and Whitehead’s company grew to meet this AI moment.

Transforming the American grid for AI-ready power

Giga Energy sits in a unique position as a vertically-integrated builder of AI infrastructure. The company blends in-house American manufacturing (nearly 4 gigawatts of equipment delivered to date) and its experience in data center development.

While startups in Silicon Valley, Austin, New York, Boston and other traditional tech hubs scramble to code the next AI tool, Giga Energy is building the hard infrastructure that AI companies need. If you think of AI as a type of tech “gold rush”, Giga Energy sells the picks and shovels that make it possible.

Manufacturing hasn’t diminished the company’s drive to put equipment to use, though. Prominent AI companies hire Giga Energy to build “turnkey sites,” brand new data centers that AI hyperscalers take over and deploy GPUs themselves. Other companies can rent rack space at data centers that Giga Energy owns and operates, a practice known as “colocation,” which can allow AI companies to bring GPUs online more quickly without needing to make the substantial upfront investments typically required for new infrastructure.

“Solving the problem of AI power isn’t just about making more transformers,” said Whitehead. “You need to find good sites, build the infrastructure, and ultimately energize something ready to handle compute loads that didn’t exist just a few years ago.”

For the founders of Giga Energy, building the next generation of energy infrastructure feels like a familiar challenge.

“Building new data center sites for AI should make the grid more resilient, not less,” said Lohstroh. We are currently operating 175 megawatts of project sites that we built ourselves, with our own equipment and have more than 500 megawatts in our pipeline right now. It’s not a question of how anymore for us, but a question of where we build next.”

Investing involves risk and your investment may lose value. Past performance gives no indication of future results. These statements do not constitute and cannot replace investment advice.