Revolut launches AI research unit behind its in-house model PRAGMA
What's the deal? Revolut has launched Revolut ResearchDealroom has a profile for this one. Try Dealroom →, a dedicated division within its AI department focused on building machine learning architecture for financial services. The unit is the engine behind PRAGMA, Revolut's proprietary foundation model, developed with NVIDIA to power risk assessment, platform operations, and product recommendations.
What's the endgame? Revolut wants to build its AI stack in-house rather than rely on third-party tools. Where rival banks build separate models for tasks like fraud and customer service, PRAGMA is a single foundation that "learns patterns across the entire customer journey," according to Revolut.
Early results: Testing on historical data showed gains over legacy baselines, the company said. Those include 2.3 times higher accuracy in identifying credit default risk, 65% more fraud cases caught with 17% greater precision in alerts, and 41% more relevant product recommendations.
Why now? Revolut argues the banking industry is hitting "a technological wall" as legacy institutions retrofit old infrastructure with off-the-shelf software. It positions its own approach as embedding intelligence directly into its financial engine.
The unit is led by Anton Repushko, who has presented findings this year at conferences including ICML in Seoul. It will work alongside academic and technology institutions.
What's the edge? Revolut points to its dataset: 80 million customers across more than 40 markets, processing billions of cross-border transactions in real time. "As the dataset grows, the models become exponentially smarter at detecting fraud, evaluating risk, and predicting user needs," the company said, calling it an advantage "that traditionally structured banks struggle to match."
"To lead the future of intelligent banking, you cannot rely on third-party blueprints," said Pavel Nesterov, head of AI at Revolut. "We have launched Revolut Research to institutionalise our 'build, don't bolt on' philosophy."
The signal: The launch reflects a wider push among fintechs to treat proprietary AI as a competitive moat rather than a bolt-on feature. By centralising research around a single foundation model, Revolut is betting that scale of data — not licensed tools — will decide who leads intelligent banking.
Read more: Tech.eu, revolut.com
Image credit: Generated with Gemini
