1M+ Users Verify for Online Loans with 3DiVi Facial Recognition

A major online lending platform with over 1 million users, faced a challenge as its customer base rapidly expanded. The company needed to speed up loan approvals while maintaining strong fraud defenses. The primary goal was to automate borrower identity verification with facial recognition, reducing the workload for underwriting staff and the risk of fraudulent applications.

The rapid growth in online applications highlighted several critical issues:

• Fraudsters frequently attempted to impersonate legitimate users during the application process.

• Manual verification was slow, error-prone, and labor-intensive.

• Staff had to manually check multiple factors, including signs of photo manipulation, matching between the face on a selfie and the passport photo and consistency between the registered address in the form and the address on the passport.

Each missed fraud attempt using fake or stolen photos and documents risked significant financial losses.

To automate the loan application submission, the company implemented 3DiVi Face SDK, enabling:

• 1:1 Face Verification – Comparing the face on a selfie with the passport photo.

• 1:N Face Identification – Searching a database of over 1 million users to detect duplicate accounts created by fraudsters.

A server-side solution was integrated with the online application form:

⓵ The user uploads a selfie alongside their passport.

⓶ The system, powered by 3DiVi Face SDK, performs:

• 1:1 face verification between selfie and passport photo.

• 1:N face search against the user database and blacklists.

⓷ If all checks pass, the application proceeds automatically. If any check fails, it is flagged for manual review.

Implementation of facial recognition technology delivered measurable benefits:

• 400+ fraudulent applications are detected monthly, where fraudsters attempt to forge selfies or passport photos.

• 95%+ accuracy in identifying fakes.

• Application processing time and costs dropped thanks to reduced staff workload.

• Verification became nearly instantaneous, virtually eliminating the risk of issuing loans using stolen or fake identities.

By integrating facial recognition, out client turned a growing risk into a competitive advantage: faster approvals, fewer fraud losses, lower operational costs and a smoother experience for legitimate users.

Interested in adding face biometrics to your project? Book a free consultation and see what computer vision can do for your business.