DeepMind Trio Behind Poker AI Now Generating Returns for Quant Hedge Funds

Three ex-DeepMind researchers, who previously developed an AI capable of defeating human poker champions, have pivoted that technology to financial markets — and the strategy is yielding significant returns. Their Prague-based AI venture, EquiLibre Technologies, has reached a $500 million valuation following a Series A funding round of undisclosed size, according to TechCrunch.

Creandum led the investment round. Although the venture capital firm did not specify the amount, Vice President Cameron Sellers told TechCrunch that this represents the largest single investment the firm has ever made in a company.

The link between poker and Wall Street lies in their suitability for reinforcement learning, an AI training method where self-learning models are rewarded for specific outcomes. Martin Schmid, CEO of EquiLibre, explained, “The advantage of trading and markets is that the scoring mechanism is straightforward: how much profit did the agent generate?”

This is not theoretical play. Partnering with quantitative firm Tower Research Capital, EquiLibre’s algorithms are currently managing billions in daily trading volume across the S&P 500 and Nasdaq. The startup reports strong performance since its launch on cryptocurrency markets in 2025, and subsequently on stock exchanges, boasting “a perfect record of zero negative months since inception,” meaning every month has ended with positive overall returns.

By targeting quantitative hedge funds, the startup operates in a sector where automation is standard, and successful innovations can be rapidly monetized. This potential made EquiLibre attractive to Creandum, according to Sellers.

“The total addressable market for trading in financial systems is among the largest globally, with countless funds generating profits that dwarf most venture-backed successes,” Sellers noted. However, he emphasized that EquiLibre explicitly identifies itself as “a lab first, not a finance firm.”

Schmid and his co-founders — CTO Rudolf Kadlec and CSO Matej Moravcik — lack traditional finance backgrounds, a fact they do not hide. “I’m not motivated by making markets more efficient,” Schmid told TechCrunch. “I’m driven by the excitement of building something entirely new, and the process is genuinely enjoyable.”

Frontier AI ventures founded by DeepMind alumni are currently highly sought after by venture capitalists. Another recent example is Ineffable Intelligence, which recently secured $1.1 billion in funding. While most of these companies are based in the U.K., notable exceptions include EquiLibre.

The founding trio of EquiLibre were visiting PhD students at Google’s first international AI research office in Edmonton, Alberta, Canada (which Alphabet closed in 2023). During their time there, they developed DeepStack, the first AI program to defeat professional players in no-limit poker, commonly known as Texas hold ’em. They also collaborated with professors who now serve on the startup’s prestigious advisory board, including Rich Sutton, who received the 2024 Turing Award for his contributions to reinforcement learning.

To establish their startup, EquiLibre’s founders decided to return to their home country, the Czech Republic. “This is where we had many collaborators, and there was a significant Czech diaspora at Google and other tech hubs,” Schmid explained. “These were our friends, so we asked them, ‘Hey, we’re moving back to Prague. Do you want to join us?’”

This approach helped EquiLibre assemble its initial team in 2022 and grow to its current staff of 25. According to Schmid, this location choice continues to offer advantages. Compared to San Francisco, “It is much easier to retain top talent here, because there isn’t a new, flashy AI trend emerging every two months.”

EquiLibre is not the only prominent AI startup in the area; BottleCap AI is located in the same building.

Nevertheless, this remains one of the region’s most notable AI companies for attracting talent. The next step is to scale its compute infrastructure, launching what is expected to be one of the largest compute clusters in Central and Eastern Europe (CEE).

While the startup did not disclose its total funding to date, Schmid mentioned previous pre-seed and seed rounds. Pre-seed backers included Credo, a CEE-focused VC firm that also invested in ElevenLabs and UiPath. According to Dealroom data, EquiLibre’s $10 million seed round was led by Blossom Capital, valuing the company at $140 million.

Sellers confirmed that the Series A valuation of $500 million marks a substantial increase. This jump coincides with a favorable shift in sentiment toward reinforcement learning (RL), particularly in trading applications. “When we started, people were skeptical,” Schmid recalled. “Now, RL is the standard. Since we began four years ago, we believe we have a head start.”

However, there is a risk that competitors could leapfrog the startup. Trading giant Jane Street, for instance, states it already utilizes RL combined with LLMs, “or whatever else is needed to train effective models.” It also claims to possess “tens of thousands of high-end GPUs,” whereas EquiLibre aims to maximize efficiency from fewer chips, “getting more from less,” according to Schmid.

Given Jane Street’s immense profitability, EquiLibre must execute its strategy carefully to achieve its goal of becoming “the AI lab in trading.” However, unlike poker, this market may not have clear losers. As Schmid stated: “This is not a winner-takes-all market.”