Smart Engines Changes Personal Data Processing Standards at Genbank
Genbank has deployed Russian AI technology from Smart Engines to automatically recognize customers’ passport data.
Documents issued to citizens of Russia and other CIS countries are processed in seconds with 99.9% accuracy. The data, however, remains within the bank’s closed IT environment and is not transferred to external clouds.
Computer vision technology has cut the time needed to transfer passport details into the financial institution’s database severalfold – employees no longer have to enter the information manually, allowing customers to be served faster. The deployed software can process 125 passports per second on a single server.
Handwritten and Printed
The system recognizes both printed and handwritten information in fractions of a second, operating directly within the bank’s local IT perimeter without relying on external APIs or GPU clusters. This meets the requirements of Russia’s Federal Law No. 152-FZ on personal data and standards established by the Bank of Russia. The software reads key details – including identification and registered-address data – from Russian passports as well as identity documents issued to citizens of Azerbaijan, Kazakhstan, Tajikistan, Uzbekistan and other CIS countries.
OCR modules, which automate text extraction from images, can be integrated across a wide range of banking processes, including fraud prevention, remote banking services and customer profile updates. They can verify document authenticity and even detect deepfakes, an increasingly relevant capability as generative AI becomes more widespread.
By deploying the new technology, Genbank has accelerated customer service across all its branches and minimized errors when entering passport data, while also making it easier for employees to work with foreign nationals.
Smart Engines is included in Russia’s domestic software registry. Its AI-based technologies are accelerating the replacement of foreign software and helping strengthen Russia’s technological sovereignty.
Domestic and International Markets
Smart Engines products are used not only in Russia but abroad as well. They are exported to Armenia, Belarus, Kazakhstan, the United Arab Emirates and Turkey. “We see significant potential for cooperation with companies in Latin American countries, as the region is showing dynamic growth in fintech and the automation of banking services,” CEO Vladimir Arlazarov said last year.
The company is now investing in promoting its technologies, including age recognition from identity documents, in Kazakhstan, Uzbekistan, the UAE, India and China. The ability to recognize documents in 100 languages and process them locally without transferring data to third parties gives the technology a competitive edge for banks across the CIS and BRICS markets, where data-sovereignty requirements are becoming more stringent.
Scaling Smart Engines
One of the first examples of the technology entering the international identity-verification market came in 2021, when the Russian Smart ID Engine was integrated into the global Argos KYC service. At the same time, its use was scaling up in Russia. In 2022, Alfa-Bank deployed the technology to recognize passport data in real time when updating customer information, while Credit Europe Bank integrated the OCR technology into its mobile app in 2023. A year later, T-Bank expanded its geographic coverage by adding recognition of Belarusian passports to speed up account opening for Belarusian citizens. These are only a few examples – Smart Engines solutions have been deployed at scale, and not only in finance. They are also used by digital financial asset operators, real estate developers and notaries. Last year, the company developed Sherlock 2o, Russia’s first multimodal AI model for combating forged documents. When verifying a document, the model analyzes 600 parameters to determine its authenticity.
“The system can be used in any scenario that requires documents to be presented – at banks, microfinance organizations, insurance companies and government agencies,” Vladimir Arlazarov said.
From Optional Feature to Core Standard
The market is evolving from basic text extraction toward intelligent document verification and fraud analytics. AI-powered document recognition has moved from an auxiliary feature to a core component of the digital front office, where what matters is not only OCR speed but also the ability to operate within a sovereign IT environment subject to stringent security requirements.
Over the next two years, the range of supported documents will expand, along with integration with biometric authentication systems. Demand for technologies capable of detecting manipulation will also grow. For the financial sector, such systems are set to become a core standard as digital sovereignty requirements increase.
