Case study: a low-code decision engine for assessing the credit risk of a retail client in Bank Pocztowy

Bank Pocztowy S.A. is a Polish retail bank. The Bank's mission is to provide customers with simple, secure financial services at a good price. Thanks to its strategic partnership with Poczta Polska S.A., it gives its Clients access to financial services through one of the widest networks of branches in Poland.

The implementation team for the financial sector of yarrl S.A. (previously in ProService Finteco and Bonair) has been cooperating with Bank Pocztowy S.A for years in the field of advanced technologies supporting credit processes. Thanks to this cooperation, Bank Pocztowy S.A. gained access to an innovative tool - a low-code decision engine, which supports analysts in creating and maintaining credit risk assessment models.

The low-code design team adapted the solution to existing processes, integrating them, among others, with the databases of the Credit Information Bureau (BIK). The key stage was the training of the Bank's analysts, who gained the opportunity to independently create and modify decision models. With drag-and-drop tools, analysts can quickly develop and modify scoring algorithms and evaluate BIK reports without the need for developers or third-party technology providers.

The proprietary low-code platform provides an extensive library of components and advanced algorithmic logic modeling capabilities, allowing analysts to create, integrate, test, and deploy decision models intuitively. This environment, aimed at people with good business knowledge, does not require advanced programming skills or deep technical knowledge, which is in line with the concept of “citizen developer” — users who can create modern IT solutions with the help of simple graphic editors.

The new technology has brought a number of benefits. Thanks to low-code tools, analysts can quickly create algorithms, which shortens the time to implement changes and increases the Bank's flexibility in responding to market needs. The decision engine also enabled a more precise assessment of credit risk and more accurate recommendations, while reducing operating costs by eliminating the need for external technical support.

Low-code decision engine for creating and maintaining algorithms for retail client risk assessment and evaluation of BIK reports, independently by bank analysts:

• integrated into the bank environment and the credit process,
• has advanced tools for designing, verifying, testing and documenting scoring models,
• integrated into the database of the Credit Information Office,
• evaluates BIK reports,
• assesses risks and provides recommendations in the credit process,
• provides a runtime and algorithm debugging module.

The process of designing scoring algorithms and evaluating BIK reports is carried out according to the philosophy of low-code - no coding.