Three former DeepMind researchers who developed an AI that beat humans at poker are now applying the same technology to stock trading, and it looks like their bet is paying off. The company’s Prague-based AI lab, EquiLibre Technologies, is currently valued at $500 million after raising an undisclosed Series A, according to TechCrunch.
The round was led by Creandum, and the VC also declined to disclose the size of the round, but vice president Cameron Sellers told TechCrunch that he confirmed it was the “largest single investment the company has ever made in a company.”
What poker and Wall Street have in common is that self-learning models are well-suited to reinforcement learning, an AI training method that is motivated by rewards. EquiLibre CEO Martin Schmid says, “The good thing about trading and markets is that the scoring is very simple: how much money did the agent make?”
This is more than just game money. In partnership with quantitative firm Tower Research Capital, Equilibre’s algorithms generate billions of dollars in daily trading volume across the S&P 500 and Nasdaq. The company claims that the agent has performed well since its deployment in the crypto market in 2025, and currently has a “perfect record of zero negative months since inception” on the stock exchange, meaning it ends each month with an overall increase in the amount invested.
By applying its AI to quantitative hedge funds, the startup enters a space where automation is the norm and, if successful, can quickly turn improvements into cash. That’s what made the startup attractive to Creandum, Sellers said.
“The total potential trading market in the financial markets is one of the largest on the planet, and there are countless funds that have generated such huge returns over the years that most venture-backed successes are dwarfed,” Sellers said. But he pointed out that EquiLibre clearly defines itself as “a research lab first, not a financial company.”
Schmidt and his two founders, CTO Rudolf Kadlec and CSO Matej Moravczyk, don’t have a financial background, and that’s not what drives them, he told TechCrunch. “We’re not doing it because we’re excited about making the market more efficient. We’re doing it because we’re all excited about building something new that has never been built before. This is a lot of fun to build,” Schmidt said.
The frontier AI potential of DeepMind alumni is an area that is also being enthusiastically pursued by VCs. A recent such example is Ineffable Intelligence, which recently raised $1.1 billion. Most of these are based in the UK, but there are notable exceptions such as EquiLibre.
In the case of EquiLibre’s founding trio, they were visiting PhD students at the Google-owned company’s first international AI research office (Alphabet will close in 2023) in Edmonton, Alberta, Canada. So they built DeepStack, the first AI program to beat professional players at no-limit poker, also known as Texas Hold’em. They also worked with professors who are now part of the startup’s distinguished advisory board, including Rich Sutton, winner of the 2024 Turing Award for his work in reinforcement learning.
To build their startup, EquiLibre’s founders decided to return to their native Czech Republic. “This is where we had a lot of the people we worked with, and there was a large Czech diaspora at Google and other places,” Schmidt said. “They were friends of ours, so we said to them, ‘Hey guys, we’re going back to Prague. Would you like to come with us?’
With this, EquiLibre built its first team in 2022, reaching its current number of 25 employees. But Schmidt says the location choice continues to pay dividends. Compared to San Francisco, “there’s not some new sexy AI thing happening every two months, so it’s much easier to keep good people here.”
EquiLibre isn’t the only AI startup gaining attention in town. BottleCap AI is located in the same building.
Still, the company is one of the most notable AI companies in the region when it comes to talent. Next, the company plans to expand its computing infrastructure and bring online what is expected to be one of the largest computing clusters in Central and Eastern Europe (CEE).
The company also declined to disclose its total funding to date, but Schmidt said it previously raised money from pre-seed backers including CEE-focused VC firm Credo, and two other funding rounds that also supported Eleven Labs and UiPath. EquiLibre’s $10 million seed round was led by Blossom Capital at a valuation of $140 million, according to Dealroom data.
The seller acknowledged that the $500 million Series A valuation was a significant inflated valuation. But that’s also what happens after the winds change in favor of reinforcement learning (RL), including trading. “When we started, people were skeptical,” Schmidt said. But now RL is the standard. “We started four years ago, so we believe we are making progress.”
Still, there is a risk that the startup could be leapfrogged by competitors. For example, trading giant Jane Street says it is already using RL, including LLM, or “whatever it takes to train good models.” And while the company claims to have “tens of thousands of high-end GPUs,” EquiLibre aims to squeeze more computing power out of fewer chips and “get more from less,” Schmidt said.
Given Jane Street’s profitability, EquiLibre will need to play its cards right to achieve its goal of being known as the “AI Lab of Trading.” But this is not poker, so there may not be any losers. “This is not a winner-take-all market,” Schmidt said.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
