PayTech Awards 2023 FINALIST: INSPIRED JV LLC

INSPIRED JV LLC is a proud finalist in the PayTech Awards 2023 for the Best User/Customer Experience Initiative – Consumer Payments category, showcasing their innovative Virtual Identification project in Payme.

In October 2021 Central Bank of Uzbekistan allowed banks and financial institutions to identify their customers virtually, on a distance. That regulation allowed market players to offer new product and services, like instant virtual cards, deposits and so on.

The same time, identification process itself is not that easy. It must, from one side, ensure perfect compliance with the law and KYC regulations; from the other side – provide perfect customer experience and convenience.

Payme had virtual identification process; but in 2022 company decided to work on its improvement.

For the identification, previously it was required to pass few steps including making selfie with passport. It was complicated, as passport is laminated, and excessive lights spoil the picture.

According to new process, if user meets the liveliness criteria for more than 65% and his photo is similar to the photo in the passport or ID card for 65% and more, then this user is considered identified, which allows Payme to trust him as a unique individual and provide an expanded range of financial and other services.

The process of determining the liveliness of the user is to shoot a 4-second video, which is checked by a special algorithm, and check it through special software that has extensive world experience in determining live human qualities from video frames.

The functionality of checking selfie photos with photos in the databases of the Ministry of Internal Affairs includes special algorithms for the main points on the face, this solution is provided by the national licensed service provider Face_id.

There is also a check of each user undergoing identification for the fact of a match according to the AML list, which is provided by the Central Bank of the Republic of Uzbekistan. This development was implemented in Payme without the participation of third-party developers. To prevent financial and other risks, Payme does not provide its services to those persons who have a match with this list.

The identification process is as simple as possible for the user. For liveness check the neural network analyzes several frames at once capturing changes on the face, thus ensuring a real person is present in front of the camera.The same algorithm uses for faces comparison between selfie photo with the person’s document.

Neural network analyzes is the key technology that gives high estimation that the user going for identification is real person rather than photo in the paper.Previously, there were many errors during identification, about 35% of all requests for identification were rejected due to an unclear photo document, since in the old process it was necessary to photograph the passport. Now there is no need to take a photo of the passport, but it is only necessary to take a selfie of the user. This significantly increased the successful passage of identification by 25%. This is about 15,000 identification requests per month. The increase in successful identification had a positive impact on the business in terms of providing a wider range of financial services to identified users, for example, an unidentified user could not pay for goods / services in the. This in turn increases financial conversion and increases MAU/DAU in general.

As required by the security of the Central Bank of the Republic of Uzbekistan, all payment organizations must comply with the rules for identifying their users and measures to reduce the risks of fraud through electronic payments.

Payme, as a fintech company, are required to take action on this requirement. Based on it, a new optimized identification technology was introduced, where the user experience is maximized and the risks of transactions by unauthorized persons are minimal.

Example: we do not allow opening a virtual visa card if the user is not fully authenticated. When identifying, we have all the data about an individual, as well as confidence in his authenticity with his documents. When opening a virtual card, visa, we are completely sure that this user is not on the prohibited lists of AML and OFAC and has reliable data on his address and other important data of an individual.