The DeepMind trio who built a poker AI are now making money for quant hedge funds

The intersection of game theory and quantitative finance has reached a new milestone as EquiLibre Technologies, a Prague-based AI laboratory, quietly ascends to a valuation exceeding $500 million. Founded by a trio of former DeepMind researchers—Martin Schmid, Matej Moravčík, and Rudolf Kadlec—the firm represents a significant migration of talent from pure academic research into the high-stakes world of algorithmic trading. The startup’s rapid appreciation in value highlights a growing trend where the architectures used to solve complex, hidden-information games are being repurposed to find signals in volatile global markets.

The founders’ pedigree is rooted in some of the most significant breakthroughs in artificial intelligence over the last decade. During their tenure at DeepMind, these researchers were instrumental in developing DeepStack and Player of Games, systems that mastered "imperfect information" environments. Unlike Chess or Go, where both players see the entire board, games like Texas Hold 'em require an agent to reason through uncertainty and intuition-based bluffing. This transition from "perfect information" games to "imperfect" ones was hailed as the final frontier for game-playing AI, proving that machines could navigate human-like ambiguity with mathematical precision.

At the technical heart of EquiLibre’s offering is the application of recursive reasoning and counterfactual regret minimization to financial time-series data. In a game of poker, a winning strategy involves not just calculating odds, but anticipating how an opponent might react to one’s own actions—a recursive loop of "I think that you think that I think." Financial markets operate under similar constraints; they are non-stationary environments where the behavior of one participant changes the landscape for others. EquiLibre’s models attempt to treat market entries and exits as moves in a massive, multi-player game of imperfect information, aiming to predict price movements where traditional statistical models often fail.

The business mechanics of EquiLibre deviate from the standard SaaS (Software-as-a-Service) model commonly seen in Silicon Valley. Rather than selling a platform to a broad range of clients, the firm is deeply embedded with quantitative hedge funds, providing the sophisticated "alpha-generating" engines that drive high-frequency and mid-frequency trading strategies. This positioning allows the startup to capture a portion of the massive upside found in the financial sector, a pivot that has likely fueled its swift climb to a half-billion-dollar valuation without the public fanfare typical of consumer AI companies.

The implications for the broader financial industry are profound, suggesting an intensifying arms race in algorithmic sophistication. As AI talent leaves top-tier labs like Google DeepMind and OpenAI to launch specialized boutique firms, the barrier to entry for smaller hedge funds rises. The success of EquiLibre signals that the next generation of market volatility may be managed—or even caused—by algorithms that treat the global economy as a complex game theory problem. Regulatory scrutiny is also expected to follow, as the line between innovative trading and market manipulation becomes harder to define when the decision-making logic is buried deep within a neural network trained on game theory.

Looking ahead, the primary question for EquiLibre is one of scalability and market saturation. While their models have proven effective in the zero-sum world of professional poker, global macroeconomics involves exogenous shocks—geopolitical crises, sudden regulatory shifts, and natural disasters—that cannot always be modeled as game moves. The industry will be watching to see if the firm can maintain its edge as similar talent-led ventures emerge, and whether their specialized approach to "imperfect information" can withstand a regime change in the global economy, such as a shift away from the low-interest-rate environment that has defined the last decade. For now, Prague is no longer just a historical hub; it is the center of a new frontier in the gamification of finance.