DeepMind Poker AI Researchers Launch Trading AI Startup
Original: The DeepMind trio who built a poker AI are now making money for quant hedge funds
Why This Matters
Demonstrates practical commercial applications of frontier AI research, validating reinforcement learning for high-stakes financial systems with measurable ROI.
Three former DeepMind researchers founded EquiLibre Technologies, a Prague-based AI lab valued at $500 million after Series A funding led by Creandum. The company applies reinforcement learning to algorithmic trading for quant hedge funds, reporting zero negative months since launching on crypto and stock markets.
EquiLibre Technologies, founded by Martin Schmid (CEO), Rudolf Kadlec (CTO), and Matej Moravcik (CSO), has adapted the reinforcement learning techniques used to build poker-playing AI for algorithmic stock trading. The three founders were visiting PhD students at DeepMind's Edmonton, Alberta research office before launching the company. In partnership with quant firm Tower Research Capital, EquiLibre's algorithms currently trade billions in daily volume across the S&P 500 and Nasdaq. The startup achieved a $500 million valuation after raising a Series A round led by Creandum, which confirmed it was the largest single investment the firm has made. EquiLibre claims a perfect track record of zero negative months since inception, having launched on crypto markets in 2025 and expanding to stock exchanges. CEO Schmid emphasized the connection between poker and markets: both are well-suited for reinforcement learning because success is measurable through clear rewards—in trading, literally how much money the AI agents generate. Schmid stated the founders are motivated by building novel AI systems rather than financial innovation, explicitly positioning EquiLibre as a research lab first, not a finance firm. The company joins other frontier AI ventures by DeepMind alumni attracting significant VC interest, including Ineffable Intelligence's recent $1.1 billion raise.