A data center power startup founded by two former creator economy entrepreneurs has closed a $27 million seed round at a $500 million valuation, as the artificial intelligence industry’s appetite for reliable electricity continues to outpace grid capacity.
From Wishlists to Watts
Leonhard Soenke and Patrice Becker launched Throne in late 2021 as a wishlist platform that let fans buy gifts for their favourite content creators. The business grew quickly. According to TechCrunch, Throne surpassed $1 million in monthly gift purchases with 50% month-over-month growth in 2021, and later took the rare step of returning early investor money it had raised via a SAFE note, choosing to run the business without outside capital.
By the time Throne had grown into a team of roughly 40 staff across offices in New York, Dallas, and Berlin, per Business Insider‘s 2023 reporting on its seed pitch, the founders were already feeling the pull of something larger. Throne had a third co-founder, Heiner Stinner, alongside Soenke and Becker. The pair handed day-to-day control to trusted long-term team members over roughly six months before turning their attention elsewhere.
Soenke told Business Insider that the creator economy had matured into a crowded space, and that he and Becker were drawn to sectors with bigger, harder problems. They moved to San Francisco, spent time with researchers at AI laboratories and major compute providers, and concluded that power supply was the sharpest constraint on AI infrastructure growth. Their new venture, TAR (Transformative American Resources), is the result. The company has since relocated its base to Austin, Texas, keeping a San Francisco office for engineering work.
TAR’s Data Center Power Model
TAR’s proposition, as reported by Forbes, centres on developing data centre campuses co-located with renewable energy and storage assets. The $27 million seed round is earmarked for deploying scalable on-site, plug-and-play power infrastructure.
The on-site approach matters for more than efficiency. According to Data Center Dynamics, TAR’s behind-the-meter systems reduce dependence on public grid infrastructure, which in turn allows data centre installations to be built farther from large urban areas where grid capacity is already strained.
Becker told Forbes that the energy mix will span solar, batteries, wind, and simple cycle gas turbines for emergencies and prolonged periods of unfavourable weather. The nameplate capacity of the system is much larger than 10 MW, with the lower operational figure reflecting fluctuations in renewable generation rather than any hard ceiling on scale.
Soenke described TAR’s role as innovating the deployment of existing technologies rather than inventing new methods of energy production. The systems are modular and designed to be installed quickly on-site. The investor behind the seed round has not been named publicly. TAR counts the shift to this kind of work as a material change in operational reality: the business involves procuring heavy equipment, developing land, and building physical infrastructure rather than shipping software updates from a city office.
Leaving Throne Behind
Soenke acknowledged that stepping back from Throne was not straightforward. ‘Our attitude is if ain’t broke, don’t fix it,’ he said, wearing a Patagonia quarter-zip at the TAR meeting that he described as nearly identical to the one he had worn throughout the Throne years, now with TAR’s logo on the chest.
Throne had grown into a meaningful business, with partnerships across major vendors and a working relationship with Amazon among its commercial ties. Letting go, Soenke said, required trusting the team left in place and a gradual handover process that stretched across several months.
The founders argue that the two businesses share an underlying logic: both required identifying a problem with genuine market pull and building fast to meet it. At Throne, that pull came from creators wanting a simple way to receive gifts from their audiences. At TAR, it is AI operators needing reliable power that the grid currently cannot guarantee at the pace they require.
The seed round provides the runway to prove the model works at scale. The question of whether TAR can deploy its modular systems fast enough to capture demand before larger incumbents adapt will determine whether a $500 million valuation at seed stage reflects ambition or early evidence.