AI on the Go: Banks Utilize Computer Vision for Card Issuance


Otkritie Bank (part of the VTB Group) has implemented artificial intelligence (AI) technology for automatic passport recognition during offsite customer service. The use of AI is expected to speed up the card issuance process during offsite services by 15-20%.

'When receiving a bank card, the client needs to present a Russian passport, and the bank representative must verify its details against the application. After the technology implementation, the process became simpler: the bank representative now only needs to photograph the main spread of the passport. The photo is sent to the server at Otkritie, where all the necessary information is automatically recognized, and the correctness of the data entered by the client when filling out the application is checked. The image processing time is less than a second,' reports the press service of Otkritie Bank.

According to the bank, the document recognition system is integrated into the bank's internal information circuit: images of documents and clients' passport data are not transmitted to external services in any form - processing is performed autonomously, ensuring confidentiality, and eliminating the risk of personal data leakage.

Rodion Martynov, a lead project manager at Cinimex, positively evaluated the solution, noting significant optimization in application processing time. By minimizing the human factor, security is enhanced, and the bank can be confident that the passport presented for card issuance has details identical to those in the application. The next step, he suggests, could be taking a photo of the passport next to the client's face and automatically comparing faces.

'Computer vision technologies that allow such analysis have been available on the market for over five years. In banks themselves, as far as is known, the client's passport can also be read using OCR (Optical Character Recognition) technologies. The solution itself is not new, but modern approaches allow it to be implemented in new banking processes, such as offsite services,' believes Rodion Martynov.

He noted that many banks use computer vision - for example, the 'pay with a smile' service.

'By analogy, banks can use face recognition technologies for registration and logging into personal accounts or mobile applications, for assessing customer emotions during service, using ATMs, and for analyzing the flow from ATM cameras. Additionally, phone number recognition is used in the banking sector for transfers via the Faster Payments System (FPS),' added Rodion Martynov.


Ask your question

Next publication
How 'Podeli' Was Created
30 November 2023 | Publication