The startup’s transition from gaming to finance relies on the premise that markets share the same structural simplicity as poker: a clear, quantifiable reward system. By partnering with Tower Research Capital, EquiLibre’s models are currently executing billions in daily volume across the S&P 500 and NASDAQ. CEO Martin Schmid reports that the algorithms have maintained a perfect record of zero negative months since their rollout on crypto markets in 2025, a performance record that helped drive the firm’s latest funding round.
In section Startups & Technology
DeepMind Alumni Turn Poker-Playing AI Into a $500 Million Hedge Fund Bet
Three former DeepMind researchers who once mastered Texas hold’em are now applying reinforcement learning to Wall Street. Their Prague-based startup, EquiLibre Technologies, has reached a $500 million valuation after a major Series A led by Creandum, signaling a shift in how automated algorithms navigate complex financial markets.
EquiLibre’s founders—Schmid, Rudolf Kadlec, and Matej Moravcik—built their reputations by creating DeepStack, the first program to defeat professional poker players. Their work in Edmonton, Alberta, alongside reinforcement learning pioneer Rich Sutton, provided the foundation for their current approach. Despite the high-stakes environment of quant trading, the team maintains a research-first identity, preferring the stability of their Prague headquarters over the frantic talent wars of Silicon Valley. With a team of 25 and plans to build one of Central Europe’s largest compute clusters, they aim to outperform industry giants like Jane Street by maximizing efficiency rather than brute-force computing power. While competitors remain aggressive, Schmid views the financial sector as a vast, non-zero-sum landscape where their early start in reinforcement learning offers a distinct competitive edge.
Comments (0)
No comments yet. Be the first!