Revolut Now Automates 90% of Customer Support With AI
Founder Nik Storonsky says one developer can now do front-end, back-end, and data science work, so a 10-person team should become two or three people.
David Senra · Competition, Intensity & Building a $115 Billion Global Bank | Nik Storonsky, RevolutRevolut, the digital bank, now handles 90% of customer support with artificial intelligence, according to founder Nik Storonsky.
The support operation once employed more than 5,000 people, and is now much smaller, Storonsky says. Language models also handle translation and cross-selling, and engineering speed is up 30%.
Storonsky says he is trying to change how he organizes people after 10 years of refining that system. He says team structure needs to be rethought, with two or three people where there were 10, because one developer can now cover front-end software, back-end software, and data science.
His next bet is a full AI employee with an email account, a Slack account, and access to internal systems. It would join calls and do work, not just chat, a shift he expects most companies will eventually make.
Questions people ask
What do AI support systems actually do?
It means software that answers routine questions, checks account details, issues refunds, and escalates harder cases to a person. At many banks and retailers, it now resolves a large share of chats without a human ever joining.
What is a full AI employee?
Storonsky describes it as AI with a company email, Slack access, calendar, and permission to use internal systems. Instead of only answering prompts, it would attend meetings, complete multi-step tasks, and follow up on its own.
How can two people do the work that used to take 10?
One developer can now use AI coding assistants to draft interfaces, write server code, test it, and analyze data. That does not remove the need for judgment, review, and product decisions, but it reduces the need for separate specialists on small projects.