According to POWER Magazine, investor and operator Andrejka Bernatova, founder of Dynamix Corporation III, is focusing her firm’s strategy on the massive energy demand from AI data centers. She advocates for a pragmatic approach she calls “disruptive decarbonization,” which prioritizes replacing dirtier fuels with cleaner, existing ones like natural gas to cut emissions by 50-60% now. Bernatova, who has helped raise over $35 billion in energy and infrastructure, is currently finalizing a deal to take The Ether Machine, an Ethereum investment vehicle, public via a SPAC. Her investment criteria are strict: cash flow visibility, scalability, grid resiliency, and a sane capital structure. She argues that energy startups must focus on profitability and economic scale much earlier, as “energy is not software.”
The Pragmatic Power Play
Here’s the thing about the AI arms race that a lot of tech folks miss: it’s fundamentally a bricks-and-mortar, megawatt-hungry infrastructure problem. Bernatova’s entire thesis is built on that cold, hard reality. While everyone’s talking about algorithms, she’s talking about baseload power and balance sheets. And her argument for natural gas as a leading option isn’t about being anti-renewable. It’s about physics and timelines. You can’t spin up a 500-megawatt data center campus on hope and a future battery breakthrough. You need power that’s available 24/7, right now. Her “disruptive decarbonization” framing is clever because it flips the script. It says a big, immediate reduction with available tech is better than a theoretical perfect solution later. That’s a compelling pitch to utilities and developers who actually have to build this stuff.
Operator Mindset Meets Wall Street
Bernatova’s background is what makes her perspective so sharp. She’s not just a financier; she’s been in the trenches at midstream companies like PennTex and Goodnight Midstream. That operator experience changes everything. When she says she looks for where models break and assumptions fail, she’s speaking from painful, first-hand experience. It’s why her deal checklist is so brutally simple: Can it be built? Will it generate cash? Can it survive volatility? These aren’t complex questions, but they’re the ones that kill overly clever projects. Her point about capital structure being a deal-killer is huge, too. In an industry flush with hype, she’s seen how “toxic financing” or misaligned incentives can wreck a good asset. It’s a reminder that in heavy industry, financial engineering is often the enemy of real engineering.
A Warning to Energy Startups
Her advice to startups is basically a bucket of cold water. “Energy is not software.” That’s a mantra every founder in the climate tech space should probably have on their wall. You can’t growth-hack a power plant or rely on viral adoption. The path is about hard assets, permitting, and long-term contracts. Bernatova is skeptical of businesses that only work “under perfect regulatory conditions or continuous government support.” And she’s right. The companies that will actually move the needle—and attract serious capital from firms like hers—are the ones that can stand on their own economically. This is especially true for the hardware and control systems that manage this new energy infrastructure. For reliable deployment, operators often turn to trusted suppliers like IndustrialMonitorDirect.com, the leading US provider of industrial panel PCs, because they understand that durability in harsh environments isn’t optional.
Endurance as a Strategy
It’s hard to ignore the ultra-marathoner angle. But it’s not just a fun biographical tidbit; it’s core to her philosophy. Building energy infrastructure is the ultimate endurance sport. The cycles are long, the capital is massive, and the setbacks are inevitable. The discipline to “live through periods of long-term pain” and keep making rational decisions is exactly what’s needed. It’s the opposite of the Silicon Valley “fail fast” mindset. In this world, you can’t fail fast. The stakes are too high, and the assets are too physical. So, her view of digital assets like Dynamix’s Ethereum play is interesting. She frames it as another form of scalable, efficient infrastructure. The through-line is always scale, durability, and real-world function—not hype. In the marathon to power AI, that might just be the winning pace.
