Project “Finn”: bunq’s case of building and improving its own GenAI application for operational excellence

For years, bunq’s been using AI to detect potentially fraudulent transactions, thanks to a transaction monitoring system built in-house; it helps dramatically (by 2.5 times) reduce false hits (as opposed to a “rule-based” approach often used by traditional banks) and is completely scalable. After emergence of ChatGPT the immense potential of this technology has become apparent to the wider population. Meanwhile, bunq’s been using AI for a variety of applications including spotting fake IDs when onboarding new users, automating its marketing efforts, and in-app personalization.

bunq’s new AI focus relates to a promising field of generative AI. In just under a year, bunq’s AI team has built a cutting-edge proprietary Large Language Model which is now used by hundreds of bunq employees daily. Applications include user support (the chatbot is now involved in over 60% user tickets), retrieving information, guiding bunq employees through processes, and much more. Ali’s session will focus on how using GenAI responsibly can facilitate efficient scaling of user support and drive operational excellence, resulting in enhanced user experience and streamlined operations.