Episode Summary
Executive Summary: Alan Cooper, co-founder and CEO of OnEnergy, traces his path from oil and gas and geothermal to building battery-backed power systems for AI data centers. He explains OnEnergy’s AI UPS, a proprietary battery/control layer that decouples grid volatility from massive data center loads, simplifies interconnection, and can also support behind-the-meter generation. The conversation centers on turning data centers from grid liabilities into assets.
Main Topics: Founding story and energy background (Priority: 5/5): Cooper describes an entrepreneurial career spanning oil and gas, completions, enhanced oil recovery, and geothermal, with early exposure to batteries and clean energy economics shaping OnEnergy’s eventual direction. Why OnEnergy was started (Priority: 5/5): A 2014 oil and gas downturn created pressure to diversify, leading Cooper and co-founder Ricardo to apply their controls and hardware experience to battery storage and power-resiliency problems. AI UPS architecture vs traditional BESS (Priority: 5/5): OnEnergy’s system operates in series as an uninterruptible power supply for data centers, buffering the load from the grid and/or onsite generation rather than acting as conventional backup-only storage. Grid, interconnection, and data center volatility (Priority: 5/5): Cooper argues large AI training loads create extreme volatility and interconnection challenges, but OnEnergy’s architecture makes the load predictable to utilities and reduces regulatory complexity. Hardware, controls, and supply chain strategy (Priority: 4/5): The company combines proprietary controls/software with a hardware stack built around bankable components, including silicon carbide inverters, batteries, switchgear, transformers, and custom integration IP. Business model and scaling strategy (Priority: 4/5): OnEnergy has evolved from project development into a data-center infrastructure business with hardware sales, recurring services, and potential asset ownership/operation partnerships. Company funding and operating philosophy (Priority: 4/5): The company largely bootstrapped for years, took only one institutional round, and emphasizes profitability, customer success, and capital discipline over venture-style rapid expansion.
Key Arguments: Data centers should be treated as grid assets, not liabilities, if their volatility is isolated and managed by power-conditioning technology. Traditional UPS systems only back up critical equipment; OnEnergy’s AI UPS backs up entire campuses and serves as the power path in series. Large AI training loads can swing 70% to 90% of demand, creating grid instability that utilities were not designed to handle. OnEnergy’s system can simplify interconnection because the grid only needs to model one standardized interface instead of many downstream loads. The company’s deep expertise in controls/software is as important as hardware and is a key differentiator in a technically complex market. Bootstrapping and disciplined capital allocation helped the company build credibility, operational know-how, and bankable infrastructure before the current AI surge. On-site generation and grid-tied deployments both benefit from the same buffer architecture, including gas turbines that prefer stable output. The company believes the AI buildout is structural, not temporary, and that centralized training campuses will remain important even as inference grows.
Data Points: 5 gigawatt deal: 5 GW - OnEnergy recently signed a deal with Crusoe to deploy AI UPS across multiple hyperscale campuses. Operating/under construction capacity: More than 3 GW - Cody says OnEnergy already has over 3 GW operating or under construction. Countries deployed: 6 countries - Cooper says the company has deployed projects across six countries. Project count: About 80 projects - OnEnergy’s software/control system has been used across roughly 80 projects. Institutional capital rounds: 1 round - Cooper says the company has taken only one institutional capital round in its history. Texas grid investment: About $400 million - Cooper says OnEnergy deployed roughly $400 million into the Texas grid. Technical validation spend: $30 million to $40 million - Cooper estimates investment in technical validation over the past two years. Company revenue share from data centers: 95% - Cooper says data center opportunities now account for the vast majority of revenue. OnEnergy scope in powered shell projects: Around 10% - He says the company is typically about 10% of a powered shell construction scope. Mexico geothermal penetration: About 8% - Cooper notes geothermal makes up about 8% of Mexico’s energy grid. U.S. oil production: Almost 12 million barrels/day - He cites the U.S. as producing nearly 12 million barrels per day, though he notes the number may be slightly out of date. Revenue decline during oil crash: Almost 90% in one quarter - A 2014 oil price war caused a severe drop in his prior business’s revenue.
Pivotal Quotes: "Data centers can actually be an important asset to the grid." — Alan Cooper: Core thesis of the interview: reframe AI infrastructure from grid burden to grid support. "Our system is fully electrically disconnecting the grid from the load." — Alan Cooper: Explaining how OnEnergy’s AI UPS differs from traditional battery storage and standard UPS products. "This is the largest deployment of capital at the highest velocity in human history." — Alan Cooper: Describing why the current AI infrastructure buildout is a major, long-term opportunity.
Implications: OnEnergy’s approach suggests AI data centers may accelerate, not strain, grid modernization if paired with buffer-layer power electronics and controls. For utilities, developers, and regulators, interconnection could become simpler and communities may see more reliable power, lower risk, and better local economics.