Crypto Exchange Coinbase Taps Machine Learning To Improve Identity Verification

Announcements, News | February 13, 2018 By:

Cryptocurrency exchange Coinbase has announced that it is now using machine learning for identity verification.

Coinbase has been adding a massive number of users every day. At its peak, Coinbase was adding 100,000 users a day. In 2017, Coinbase began the year with a little over 5,000,000 users and is currently at about 13,500,000 users. As a result, many of its customers have had a poor experience with identity verification, payments, and customer support.

To fix these problems, Coinbase replaced its identity verification system with a newer one which makes use of machine learning. It also integrated an additional ID verification vendor to help with increased loads during periods of peak demand.

Coinbase said that machine learning has helped to increase the success rate of identity verification to 90%. The remaining 10% is because of the ID provided by the users not being clear or inappropriate.

“10% of ID verifications still fail, resulting in customers needing to contact the Coinbase support team,” the exchange said. “The top reason customers are unable to verify an ID is a cropped, blurry or low-quality photo.”

Coinbase said the company is continuing to invest resources in improving its identity verification process. As part of this effort, the company recently created a dedicated engineering team focused on building out its internal identity services.