EquiLibre Technologies raises Series A at more than $500 million valuation to buy AI compute

EquiLibre Technologies, a Prague startup founded by former DeepMind researchers, has closed a Series A that reportedly values it at more than $500 million (about 438 million euros).

The round was led by Creandum, whose partner Cameron Sellers called it the venture firm’s largest single investment in one company. The amount raised was not disclosed. The valuation is up sharply from the $140 million at which EquiLibre raised a $10 million seed from Blossom Capital.

Why it matters: the three founders — chief executive Martin Schmid, chief technology officer Rudolf Kadlec and chief science officer Matej Moravcik — are betting that the same reinforcement-learning approach that beat professionals at no-limit Texas Hold’em can win in financial markets. Reinforcement learning is a technique in which an AI system learns by trial and error against a reward signal rather than from labeled examples. At DeepMind’s Edmonton office the trio built DeepStack, the first AI system to beat professional players at heads-up no-limit poker.

EquiLibre now applies that method to trading. In partnership with quantitative firm Tower Research Capital, its agents handle billions of dollars daily across the S&P 500 and Nasdaq, and the company says it has had zero negative months since inception, with crypto trading added in 2025. Those are the company’s own figures and have not been independently verified.

“The question is no longer whether this approach works. It’s how big it can get,” Schmid said. “We’ve proven the technology in the world’s biggest and most liquid markets.”

The 25-person company plans to spend most of the new capital on compute, aiming to build one of the region’s largest clusters. A lean team raising at a nine-figure valuation chiefly to buy chips shows how much of the AI trading pitch now rests on scaling infrastructure rather than headcount. Whether a strategy that thrives at billions in daily volume holds up as that volume climbs is the open question.