Revolut CEO Nik Storonsky reveals plans to build proprietary AI models and data centres

Revolut Builds Proprietary AI Models on Transaction Data

Revolut CEO Nik Storonsky revealed the company is training its own AI models and constructing dedicated data centres to power its long-term growth strategy1

1

. Speaking at the Wave by Vento event in Turin, Storonsky disclosed that Revolut processes 30 to 40 million transactions daily, creating a unique dataset for building proprietary AI models using techniques similar to large language models2

2

. The fintech giant already operates PRAGMA, a proprietary AI model trained on 40 billion transactions, app interactions, and financial events from 25 million users across 111 countries2

2

. This model handles critical decisions around fraud detection, credit risk assessment, and customer retention.

Revolut's AI-Driven Strategy Targets Global Tech Status

Nik Storonsky outlined an ambitious vision to transform Revolut beyond banking into a global tech company comparable to Meta, Google, or OpenAI, but built from Europe1

1

. The company is developing both foundation models trained on transaction data and its own large language models1

1

. Revolut plans to deploy AI agents capable of shopping for customers using secure card details, while also providing financial advice on spending, stocks, and funds1

1

. The platform will embrace an open AI ecosystem, allowing third-party agents to connect through MCP, a standard protocol for AI tools, with every action requiring customer authentication1

1

. Storonsky emphasized he would rather keep the platform open than miss features by maintaining a closed system1

1

. In July, Revolut partnered with OpenAI to bundle ChatGPT Go into its subscriptions1

1

.

Data Centres and European AI Infrastructure

Revolut is constructing its own data centres specifically to train these AI models, a significant infrastructure investment that sets it apart from competitors relying on third-party cloud services1

1

. Former Vodafone chief executive Vittorio Colao noted at the same event that Europe currently has less than 2 gigawatts of AI computing capacity, highlighting the infrastructure gap Revolut aims to address1

1

. Storonsky clarified the goal is not to compete directly with American tech companies but to achieve similar scale while maintaining European roots1

1

.

U.S. Expansion and Banking Licenses

Revolut received conditional approval from the Office of the Comptroller of the Currency for its U.S. national bank charter in September, moving closer to launching American banking operations2

2

. Storonsky stated that once Revolut secures a full banking charter in the U.S., the company plans to roll out credit cards and loans to dominate the U.S. market2

2

. If pursuing a U.S. IPO, he wants the primary listing in the United States due to the deep liquidity of American markets2

2

. The company now holds six full banking licenses worldwide, including the United Kingdom, France, Australia, Lithuania, Mexico, and most recently Colombia2

2

. Revolut was valued at $115 billion in a secondary share sale this year, making it Europe's most valuable startup and worth more than established banks like Barclays or Societe Generale2

2

.