VTB has launched a machine learning (ML) model management system, a new product of the Artificial Intelligence Factory.
Vadim Kulik, Deputy President and Chairman of the VTB Management Board, said:
“VTB’s priorities are the quality and convenience of services for customers and ensuring the security of data and user funds. As part of the digital transformation program, VTB focuses on the introduction of artificial intelligence, including for the continuous improvement of customer services and countering fraud attempts. The model management system, which marks a new step in this development, is an IT platform for unifying and simplifying the implementation of artificial intelligence technologies in the bank’s business processes. The system will allow all participants in the creation of new artificial intelligence technologies to work in a single space, exchange information, and differentiate and control all the tools and data necessary for the creation of new machine learning models.”
As part of its digital transformation, the model management system is a completely domestic development focusing on Russian software and using open-source software. The model management system is functionally superior to foreign equivalents, as it implements a prototype MLOps conveyor for the rapid introduction of models into commercial operation. The MLOps concept, adapted to the tasks of integration and implementation of artificial intelligence, is an evolution of the DevOps methodology. It is widely used to improve the speed, efficiency, and quality of team development and the implementation of any information systems.