EquiLibre raises Series A above $500m to scale RL trading agents
EquiLibre Technologies, a Prague-headquartered AI research lab founded by three former DeepMind researchers, has closed its Series A funding round at a valuation exceeding $500 million. The round was led by Creandum, the Stockholm-based venture firm, which described the deal as its largest single investment to date. The company will use the proceeds primarily to expand its compute infrastructure, with a new cluster described as one of the largest in Central and Eastern Europe, and to grow its research and engineering headcount in Prague.
The firm was established in 2022 by the team behind DeepStack, the reinforcement learning system that became the first AI to defeat professional players at no-limit poker under standard conditions. EquiLibre initially proved its trading technology in cryptocurrency markets before transitioning to US equities. The company says its agents went live on traditional markets in early 2025 and now trade billions of dollars in daily volume across S&P 500 and NASDAQ instruments. It claims to have recorded zero negative months since inception, though that figure is unverified by an independent third party.
The technology and team
EquiLibre's approach centres on reinforcement learning agents trained through self-play and market feedback rather than supervised learning on historical data. The team draws from DeepMind, Google, Jane Street, G-Research and Optiver, and counts Richard Sutton, the Turing Award laureate widely credited as a founding figure of modern reinforcement learning, as an early backer and advisory board member. One of the authors of DeepBlue, the IBM system that defeated world chess champion Garry Kasparov in 1997, also sits on the advisory board.
Martin Schmid, co-founder of EquiLibre, said: "Trading is one of the few fields where technology is the entire game. There's no sales cycle, and no marketing spend can rescue a weak product. The market is the judge, and the verdict updates every millisecond."
Cameron Sellers, Vice President at Creandum, said the investment reflects confidence in the technology's scaling potential, adding that EquiLibre was "picking a domain where the feedback loop is brutal and honest, and letting the technology speak for itself."
Market context and competitive positioning
Algorithmic and quantitative trading has been dominated for years by established systematic hedge funds and proprietary trading firms. Renaissance Technologies, Two Sigma, Citadel Securities and others have built vast technology and data advantages over decades. The emergence of deep reinforcement learning as a viable approach to live market execution represents a meaningful shift in how that competitive landscape may evolve, though barriers to entry remain high given the capital, regulatory licensing, and latency infrastructure required.
Several well-funded startups are pursuing RL-based trading strategies, but few have disclosed live production deployments at this scale. EquiLibre's claimed track record, if independently verified, would represent a meaningful data point in the debate about whether frontier AI research methods can be operationalised in high-frequency financial environments.
The EU AI Act's classification framework is relevant here. AI systems making or materially influencing financial decisions may be subject to high-risk obligations under Annex III, requiring documentation, explainability standards and human oversight provisions. EquiLibre has not disclosed how it addresses these requirements, which will become enforceable across EU member states as the Act's phased timeline progresses. As a Prague-based firm operating in EU-regulated markets, compliance posture will be an area to watch as the company scales.