UNSA uses AI to cut merchant onboarding time to eight days

UNSA, a European payment service provider with over 90,000 merchants, has reduced its average merchant onboarding time from 46 to 47 days to eight days over the last 18 months. This nearly 90% improvement in speed was achieved by breaking down manual Know Your Customer processes into smaller steps and using AI to extract data from documents such as the German commercial register extract. A specific AI tool extracts the Handelsregister-ID with audit-safe logging to verify which instance of the ID was used if multiple appear on a document. UNSA categorizes merchants into low, medium, and high-risk groups. For low-risk cases, the company aims to remove humans from the loop entirely, relying on AI for data extraction and sampling controls, which allows staff to focus on complex high-risk cases.

Beyond onboarding, UNSA has deployed AI to streamline other operational areas. In its debt collection division, an AI-powered WhatsApp agent allows customers to negotiate installment plans via text in 37 or 38 languages. This tool has increased the number of installment agreements closed per week by 30% to 35%, though internal staff currently still manually approve external communications. Additionally, AI has reduced the time spent on adverse media screening from an average of 15 minutes per hit to under three minutes, a nearly 80% improvement. To drive internal adoption, UNSA mandated that all executives in Jakob von Ingelheim’s department build their own AI agent as a performance goal. Over half of the nine executives in his direct team have already completed this objective, following an internal hackathon where employees built approximately 107 to 108 apps and nearly 60 AI agents.