Three researchers who helped create DeepStack – the first artificial‑intelligence program to defeat pro players at no‑limit Texas hold ’em – have shifted their expertise from poker tables to Wall Street. Their new venture, EquiLibre Technologies, operates out of Prague and just secured a Series A financing round led by Creandum that placed the startup’s valuation at $500 million, the venture firm’s largest single investment to date.
EquiLibre’s core product is a suite of reinforcement‑learning agents that learn by maximizing a simple reward: profit. “The nice thing about trading and markets is that the scoring is super simple: how much money did the agent make?” CEO Martin Schmid explained. Partnering with quant firm Tower Research Capital, the algorithms now handle billions of dollars in daily volume across the S&P 500 and Nasdaq, after an earlier rollout on crypto markets in 2025.
The company claims a perfect record of zero negative months since its inception, meaning each month ends with a net gain. While the startup calls itself “a lab first, not a finance firm,” the financial upside is evident. The venture’s backers see the total addressable market for automated trading as one of the largest on the planet, a sentiment echoed by Creandum’s vice president Cameron Sellers, who highlighted the dramatic jump from EquiLibre’s $10 million seed round to its current $500 million valuation.
EquiLibre’s leadership team—CEO Martin Schmid, CTO Rudolf Kadlec and CSO Matej Moravcik—have no formal finance background. Their motivation, Schmid says, is “building new things that have never been built before” and enjoying the challenge. The trio originally met while visiting DeepMind’s first international AI research office in Edmonton, Alberta, where they built DeepStack. Their advisory board now includes reinforcement‑learning pioneer Rich Sutton, a 2024 Turing Award recipient.
Choosing to base the company in Czechia allowed the founders to tap a network of former Google and DeepMind engineers and avoid the talent churn of Silicon Valley. Since 2022 the lab has grown to 25 employees, sharing a building with other AI ventures such as BottleCap AI. The next milestone is scaling compute infrastructure to create one of Central and Eastern Europe’s largest AI clusters, a move aimed at extracting more performance from fewer GPUs.
EquiLibre’s approach arrives as reinforcement learning becomes mainstream in quantitative finance. Competitors like Jane Street already blend RL with large‑language models and operate “tens of thousands of high‑end GPUs.” Schmid acknowledges the risk but stresses that the market is not a winner‑takes‑all scenario. “We’re trying to get more from less,” he said, underscoring the startup’s focus on efficiency as it seeks to cement its reputation as the AI lab that reshapes trading.
Dieser Artikel wurde mit Unterstützung von KI verfasst.
News Factory APP - agentische News für besseres SEO & AEO.