Blab AI
Former DeepMind Poker AI Team Raises $500 Million to Trade Stocks and Crypto
Three ex‑DeepMind researchers who built a poker‑winning reinforcement‑learning agent have pivoted that technology to the world of finance. In Prague, EquiLibre Technologies announced a Series A that values the company at $500 million, a figure reported by TechCrunch on 30 June 2026.
EquiLibre’s founders first proved that reinforcement learning (RL) could outplay professional poker players. The same framework that taught a bot to bluff and fold now drives automated trading. The firm partners with Tower Research Capital, which executes billions of dollars of daily volume on the S&P 500 and NASDAQ. EquiLibre claims its agents have delivered stable returns since launching in the crypto market in 2025 and have posted zero‑loss months on stock exchanges.
The move from cards to markets is natural for RL. In both arenas an agent receives a reward – a win in poker or a profit in trading – and learns to maximize that signal through repeated interaction. CEO Martin Schmid says the evaluation metric in finance is straightforward: “how much money did the agent earn?” He argues that this clarity makes financial markets an attractive target for reinforcement learning.
EquiLibre’s valuation history shows growing confidence in frontier AI for finance. A pre‑seed round backed by Credo and a seed round led by Blossom Capital set the firm’s value at roughly $140 million. The latest Series A, headed by Creandum, is the largest single investment the company has received. Vice‑president Cameron Sellers notes that the new valuation reflects the rapid expansion of the RL field in trading.
The company presents itself primarily as a research laboratory rather than a conventional financial firm. Schmid emphasizes that the team’s goal is “to build new things that haven’t existed before,” not to simply increase market efficiency. Co‑founders CTO Rudolf Kadlec and CSO Matej Moravcik share that focus on experimentation and creative problem‑solving.
EquiLibre relocated its headquarters to Prague after a year of development. The founders cited the local IT community and lower talent attrition compared with Western tech hubs as reasons for the move. The lab shares a building with BottleCap AI, another Czech AI startup, underscoring the region’s growing concentration of AI talent.
Looking ahead, EquiLibre plans to scale its computing infrastructure. The team aims to launch one of the largest computing clusters in Central and Eastern Europe, a move that would support more complex RL models and higher trading volumes.
The partnership with Tower Research Capital demonstrates that EquiLibre’s agents are already operating at scale. Tower Research, a well‑known high‑frequency trading firm, provides the market access and liquidity necessary for large‑volume execution. The collaboration suggests that EquiLibre’s RL agents can handle the speed and volume demands of modern electronic markets.
EquiLibre’s success also reflects broader trends in the AI‑in‑finance space. Reinforcement learning has moved from academic research to production systems in hedge funds and proprietary trading desks. Investors are increasingly allocating capital to companies that combine advanced ML techniques with real‑world financial applications.
The Series A round signals that venture capital remains interested in AI startups tackling complex, high‑value problems. Creandum’s participation, along with the firm’s earlier seed backing, indicates a sustained belief that RL can deliver consistent, profitable trading strategies.
EquiLibre’s journey from poker to markets illustrates how a single AI technique can be repurposed across domains. The firm’s focus on research, its partnership with a leading trading house, and its recent valuation all point to a growing niche where AI, reinforcement learning, and finance intersect.
In summary, EquiLibre Technologies has secured a $500 million valuation after a Series A led by Creandum, following earlier funding rounds that valued the company at $140 million. The company’s RL agents trade billions of dollars daily on major U.S. exchanges, have a record of zero‑loss months, and are backed by Tower Research Capital. EquiLibre plans to expand its computing capacity and continue developing novel RL models, positioning itself as a laboratory that pushes the boundaries of AI in financial markets.
EquiLibre’s founders first proved that reinforcement learning (RL) could outplay professional poker players. The same framework that taught a bot to bluff and fold now drives automated trading. The firm partners with Tower Research Capital, which executes billions of dollars of daily volume on the S&P 500 and NASDAQ. EquiLibre claims its agents have delivered stable returns since launching in the crypto market in 2025 and have posted zero‑loss months on stock exchanges.
The move from cards to markets is natural for RL. In both arenas an agent receives a reward – a win in poker or a profit in trading – and learns to maximize that signal through repeated interaction. CEO Martin Schmid says the evaluation metric in finance is straightforward: “how much money did the agent earn?” He argues that this clarity makes financial markets an attractive target for reinforcement learning.
EquiLibre’s valuation history shows growing confidence in frontier AI for finance. A pre‑seed round backed by Credo and a seed round led by Blossom Capital set the firm’s value at roughly $140 million. The latest Series A, headed by Creandum, is the largest single investment the company has received. Vice‑president Cameron Sellers notes that the new valuation reflects the rapid expansion of the RL field in trading.
The company presents itself primarily as a research laboratory rather than a conventional financial firm. Schmid emphasizes that the team’s goal is “to build new things that haven’t existed before,” not to simply increase market efficiency. Co‑founders CTO Rudolf Kadlec and CSO Matej Moravcik share that focus on experimentation and creative problem‑solving.
EquiLibre relocated its headquarters to Prague after a year of development. The founders cited the local IT community and lower talent attrition compared with Western tech hubs as reasons for the move. The lab shares a building with BottleCap AI, another Czech AI startup, underscoring the region’s growing concentration of AI talent.
Looking ahead, EquiLibre plans to scale its computing infrastructure. The team aims to launch one of the largest computing clusters in Central and Eastern Europe, a move that would support more complex RL models and higher trading volumes.
The partnership with Tower Research Capital demonstrates that EquiLibre’s agents are already operating at scale. Tower Research, a well‑known high‑frequency trading firm, provides the market access and liquidity necessary for large‑volume execution. The collaboration suggests that EquiLibre’s RL agents can handle the speed and volume demands of modern electronic markets.
EquiLibre’s success also reflects broader trends in the AI‑in‑finance space. Reinforcement learning has moved from academic research to production systems in hedge funds and proprietary trading desks. Investors are increasingly allocating capital to companies that combine advanced ML techniques with real‑world financial applications.
The Series A round signals that venture capital remains interested in AI startups tackling complex, high‑value problems. Creandum’s participation, along with the firm’s earlier seed backing, indicates a sustained belief that RL can deliver consistent, profitable trading strategies.
EquiLibre’s journey from poker to markets illustrates how a single AI technique can be repurposed across domains. The firm’s focus on research, its partnership with a leading trading house, and its recent valuation all point to a growing niche where AI, reinforcement learning, and finance intersect.
In summary, EquiLibre Technologies has secured a $500 million valuation after a Series A led by Creandum, following earlier funding rounds that valued the company at $140 million. The company’s RL agents trade billions of dollars daily on major U.S. exchanges, have a record of zero‑loss months, and are backed by Tower Research Capital. EquiLibre plans to expand its computing capacity and continue developing novel RL models, positioning itself as a laboratory that pushes the boundaries of AI in financial markets.
