Sber unveils a large-scale graphics platform to optimize data-driven lending and operational efficiency

No time to read?
Get a summary

Sber has rolled out its own advanced graphics platform designed to boost the efficiency of business process management that relies on large-scale data. This strategic move was announced by the bank’s press service and signals a shift toward more data-driven decision making across its operations.

The platform is built from eight specialized services that work together to unlock insights and automate tasks. These components include graph lab, graph calculations, a comprehensive graph of client connections (EGKS), online graph access, graph visualization, a graph database for non-standard solutions, graph neural networks (GNN), and a prediction tool for connections. Each service contributes to a cohesive ecosystem that can be tailored to specific business challenges while maintaining seamless interoperability.

In practical terms, the graphical platform integrates into the lending workflow for both corporate and personal loans. This integration enables faster, more balanced, and more effective lending decisions and parameter adjustments, all based on up-to-date data relationships and network analyses. The platform is already actively supporting many automated processes within Sber, enhancing speed and consistency across credit operations and risk assessment. This capability aligns with a broader strategy to embed data science into core financial processes, helping the institution respond more nimbly to market changes and customer needs.

All services within the platform share a unified architecture, designed to interact smoothly with one another and be assembled in configurations that fit the exact business task at hand. Central to the platform is the use of open source Hadoop-based tooling complemented by Sber’s own high-performance FastGraph database, which together provide scalable storage, rapid graph processing, and robust analytics. This combination supports complex network analyses, real-time graph exploration, and the generation of actionable insights that can guide lending policies, customer segmentation, and operational optimization across diverse business units.

No time to read?
Get a summary
Previous Article

A Harsh Method of Motiviation at a Hong Kong Company Under Scrutiny

Next Article

Belarus Leader Sets Stage for Pre-Summit Talks with Putin