Biometric onboarding module for a mobile wallet app

NNTC has successfully launched a biometric onboarding module for a mobile wallet app in Oman, leveraging video analytics to help people benefit from financial services without visiting a branch office.

The project was initiated by the customer’s Head of Retail Banking. The bank chose NNTC to design fast and safe authentication of wallet users, as our facial recognition engine proved to be the most accurate and efficient and won the contract over competitors. NNTC’s facial recognition accelerates onboarding so that clients don’t even have to personally come to a bank and handle papers. The solution has never been as relevant as now, in the time of lockdown.

The app is a mobile wallet, where you can add a card issued by any bank, make a payment, transfer money to other wallets, split accounts, and perform other necessary financial transactions. It is available to the general public in official app stores. Data about client transactions and actions are protected, while the bank receives information about the interests and preferences of users to make special offers, thus converting app users into bank clients.

“User experience should be flawless, from the onboarding process to every feature. At the same time, security can’t be compromised. We expect the app to become the leader among mobile wallets in Oman,” comments Dmitry Doshaniy, General Manager, NNTC.

Smooth conversion of “user” into “client” requires maximum user comfort and process acceleration. To achieve this, the bank leveraged video analytics and artificial intelligence by NNTC. The solution was integrated into the existing mobile wallet app with help from the app support and development team.

New users make a couple of simple moves and start using the wallet. They copy the usual registration flow with the AI-enhancements :

Downloading the application

Taking a selfie. There are several options to verify that the camera is capturing a real person, but not a photo or a random video from the Internet. The app kindly asks a new user to raise the eyebrows, smile, blink, move the head sideways, move the camera to see the face from a varying distance. App’s facial recognition is based on VisionLabs algorithms used for all iFalcon solutions. These algorithms are numbers 1 and 2 worldwide in the biggest competition for face recognition by NIST of the US.

Taking photos of identity documents. The Optical Character Recognition software (OCR) recognizes the document, extracts the basic data, and compares the photo with the person’s face. If the picture of the face matches with the ID photo with the desired accuracy, a personal account will be opened automatically and a user will be notified. The bank can configure the minimum accuracy required to approve user registration. Document recognition runs on Smart Engines solution, winner of DIBCO 2017 (international Document Image Binarization Competition). The application has been penetration-tested and proved to be safe and secure.

Entering additional information (email, contact phone, address)

Thus, a user undergoes identity verification directly via a mobile device, transferring the necessary data to the bank. This accelerates user onboarding, since users don’t waste time traveling to the bank and waiting in long lines, but spend three minutes only. The bank received a new user-friendly tool to attract new clients and analyze the target audience for marketing campaign planning.

  • Customer: Bank in the Middle East
  • Year: 2020
  • Category: Video Analytics

Tell us about your business

Please fill out the form and our experts will come back with suggestions for solving them

Name *
Phone*
Subject*
accept