Duna, a two-year-old fintech founded by ex-Stripe executives, raised €30M in Series A funding led by Google parent Alphabet's CapitalG. The round included Index Ventures, Puzzle Ventures, and a notable roster of payments executives: Stripe COO Michael Coogan, former Stripe CTO David Singleton, ex-COO Claire Hughes Johnson, and Adyen executives Mariëtte Swart and Ethan Tandowsky.
The Amsterdam-based startup builds AI-native tools for business identity verification (KYB) and AML compliance. Its platform connects to 210+ registries across seven languages, offering 20+ verification modules that customers including Plaid, Adyen, Bol, and SVEA Bank use to onboard companies. Duna claims its system delivers 10.6x faster onboarding and 4.8x productivity gains compared to manual processes.
The market pain is real. European banks spend 10-20% of total expenses on compliance, with Dutch banks alone employing roughly 13,000 people in compliance roles, half focused on business customers. The manual verification process causes 30-40% of B2B signups to abandon onboarding entirely.
Duna's pitch: create a "digital passport" network where verified business data can be reused across platforms, cutting fraud while reducing compliance friction. Founders Duco van Lanschot and David Schreiber, both Stripe alumni, argue AI is accelerating fraud velocity, making automated compliance a revenue driver rather than just a cost center.
What's interesting: competitors from both Stripe and Adyen backed the round. That cross-company investment suggests the addressable market for business identity verification remains largely unsolved, particularly for smaller banks and fintechs without enterprise-scale compliance teams.
The round brings Duna's total funding above €40M. The company targets what TechCrunch calls the "long tail" of financial institutions that lack resources to build verification infrastructure in-house.
Worth noting: this is fundamentally a workflow automation play dressed in AI language. The platform ingests data from government registries and applies pattern matching. Whether that's genuinely "AI-native" or effective use of APIs and data standardization is debatable. The results matter more than the label.