Three former DeepMind researchers who once built a poker-winning AI are now applying the same machine-learning approach to financial markets. Their Prague-based startup, EquiLibre Technologies, has reached a $500 million valuation after a new Series A round.
The investment was led by Creandum, which described it as the firm's largest single commitment to date. The company's core method is reinforcement learning, a training model that rewards systems for better outcomes and makes trading a natural fit for experimentation.
EquiLibre says its algorithms are already active across major U.S. equity markets, including the S&P 500 and NASDAQ, through a partnership with Tower Research Capital. The startup also reports strong performance in crypto and stock trading, while emphasizing that it sees itself first as an AI lab rather than a traditional finance company.
The company was founded by Martin Schmid, Rudolf Kadlec and Matej Moravcik, who previously helped create DeepStack, an early poker AI developed during their time in Edmonton, Canada. Today, the team has grown to 25 people in Prague, where the founders say the local talent base has helped them build with focus and continuity.
EquiLibre now plans to expand its computing infrastructure and aims to operate one of the largest compute clusters in Central and Eastern Europe. As reinforcement learning becomes increasingly central to advanced AI systems, the company's path suggests a future where research labs and markets may evolve even more closely together.