People don't believe me when I tell them that fraud is by far the hardest problem for making a good on-ramp UX.

father, meditator and (freelance miniapp builder and consultant). this month working for $runner

You have to have 1) data (which a startup doesn't have yet) and 2) world-class fraud engineering.

Everyone is from somewhere. Cofounder and CEO of Pinata. https://www.pinata.cloud/

1. As much data as possible about the user — both KYC, payment method and then other signals

2. Historical users + payment methods fed into a machine learning model that can assess the likelihood of fraud for new users (this is the hard part for startups)