EquiLibre Technologies Hits $500M Valuation

A trio of former DeepMind researchers, renowned for developing an artificial intelligence capable of mastering complex poker strategies, has successfully translated their expertise into the high-stakes world of quantitative finance. Their venture, EquiLibre Technologies, a Prague-based AI laboratory, has achieved a significant milestone, reaching a $500 million valuation following a substantial Series A funding round led by Creandum. While the precise investment sum remains undisclosed, Creandum vice president Cameron Sellers confirmed it represents the firm’s largest single investment to date.

The core technology underpinning EquiLibre’s success lies in reinforcement learning, an AI training methodology that leverages reward-based systems for self-improvement. This approach proves particularly effective in environments characterized by clear scoring mechanisms, such as financial markets. As EquiLibre CEO Martin Schmid explained, the simplicity of measuring profit in trading makes it an ideal domain for their AI agents. The firm has reportedly partnered with the quantitative trading firm Tower Research Capital, deploying its algorithms to manage billions in daily trading volume across major indices like the S&P 500 and Nasdaq.

EquiLibre’s AI agents have demonstrated a remarkable track record since their initial deployment in cryptocurrency markets in 2025, and subsequently in stock exchanges. The company boasts a "perfect record of zero negative months since inception," indicating consistent positive returns. This performance in a sector where automation is already prevalent and incremental improvements can yield substantial financial gains made EquiLibre an attractive proposition for investors like Creandum. The sheer scale of the financial markets, estimated to be one of the largest addressable markets globally, offers immense profit potential, far exceeding typical venture-backed successes.

Despite the financial focus, Schmid and his co-founders, CTO Rudolf Kadlec and CSO Matej Moravcik, emphasize their primary motivation is the pursuit of novel AI development rather than a passion for market efficiency. Their background is rooted in cutting-edge AI research, not finance. This distinction is crucial, as EquiLibre positions itself as a research laboratory first and foremost. Their previous work at DeepMind’s international AI research office in Edmonton, Canada, culminated in the creation of DeepStack, a pioneering AI that defeated professional poker players. They also collaborated with influential academics, including Turing Award recipient Rich Sutton, a key figure in reinforcement learning and now part of EquiLibre’s advisory board.

The decision to establish EquiLibre in their home country, the Czech Republic, has proven strategic. By leveraging a network of former colleagues and a strong local talent pool, the company rapidly assembled its initial team in 2022 and has grown to 25 employees. Schmid notes that operating outside the hyper-competitive San Francisco AI scene allows for better talent retention. While other AI startups, such as BottleCap AI, are also present in the region, EquiLibre stands out for its ambitious plans to develop one of Central and Eastern Europe’s largest compute clusters. Previously, EquiLibre secured a $10 million seed round led by Blossom Capital at a $140 million valuation, with earlier support from CEE-focused VC Credo.

The significant valuation jump reflects a broader market shift favoring reinforcement learning applications, particularly in trading. While initial skepticism existed, RL is now widely recognized as a standard approach. EquiLibre believes its early start four years ago provides a competitive edge. However, the firm faces formidable competition, including trading giants like Jane Street, which reportedly employs RL with large language models and possesses extensive GPU infrastructure. EquiLibre’s strategy of maximizing compute efficiency from fewer resources will be critical as it aims to establish itself as a leading AI lab in the trading domain.