How Revolut runs AI at scale
About this episode
Nikolay Donets discusses the challenges and strategies of running AI at scale at Revolut.
Nikolay Donets, Head of Machine Learning Engineering at Revolut, on what it takes to run AI across more than 70 million customers, 200+ products, and 40+ countries - and why the hard part is no longer the model but the control plane around it: one gateway, a use-case-based governance layer, fallback chains, cost controls, and mandatory human oversight. Recorded at RAAIS 2026. Chapters: 0:00 Intro - Revolut's AI at scale 1:24 The problem: classical ML and three libraries 2:54 The 2022 shift to API-served models 4:25 Four internal groups, four sets of needs 9:39 The decision: govern the use case, not the model 10:54 One central gateway vs. distributed libraries 14:10 Performance monitoring and drift detection 17:33 Lesson: fallback chains and the silently-dead model 20:02 Lesson: frontier vs. non-frontier cost (up to 8x) 20:48 Lesson: the platform is the org chart 22:59 Case study: from Rita to AIR 26:40 Voice support at scale 28:21 AIR, the in-app assistant 30:25 Q&A: human oversight, hallucinations, AI as judge
People in this episode
Host: Nathan Benaich
Guest: Nikolay Donets
Topics covered
- AI at scale
- Machine Learning
- Governance
- Cost control
- Human oversight
- Performance monitoring
Keywords
- Revolut
- AI
- Machine Learning Engineering
- governance
- cost controls
- human oversight
- performance monitoring
Mentioned in this episode
Organizations: Revolut
More episodes of Air Street Press
- The case for intelligence beyond language
- Europe’s air defense gap
- From driving the world to dreaming it
- Turning compute into intelligence
- Accelerating science and medicine with collaborative agents
- Beyond hill climbing: the path to superhuman scientific discovery
Explore listener stats, chart rankings, contacts and more on the Air Street Press podcast page.