PSB builds a Unified Data Warehouse on Arenadata DB

2024: Arenadata DB Unified Data Warehouse

On July 24, 2024, Arenadata announced that PSB Bank, as part of its strategic development, is implementing a project to create a Unified Data Warehouse (ECD) and the Big Data platform, one of the key technological components of which is the distributed DBMS (ADB) from the Russian vendor Arenadata. As of July 2024, the work carried out within its framework made it possible to optimize the quality of client data of individuals by 56.1%.

According to the company, Arenadata DB was selected based on the results of an analysis conducted during technical approvals. Comparative testing was performed:

• on test benches with comparable equipment;
• on a single set of input data specific to the banking sector (specially developed data model);
• by executing a plurality of analytical queries simulating a typical load for data stores.

PSB specialists compared the performance of Arenadata DB and other DBMS already used in the bank for similar tasks. According to the results of testing, Arenadata DB showed optimal performance. In addition, Arenadata DB is included in the ERRP of the Ministry of Digital Development of the Russian Federation and is suitable for the tasks of building analytical data warehouses.

{{quote 'author
= Ekaterina Varlamova, CDO, Director of the Data Management Department, PSB' Our project is aimed at maintaining the bank's strategic objectives both in terms of business efficiency and in terms of data reliability and security. When choosing technologies for creating a single storage facility, we were guided primarily by the directive of the Ministry of Digital Development of the Russian Federation on import substitution. Arenadata DB not only meets the requirements of the import substitution program as a domestic product, but is also a competitive solution in the field of big data processing and storage. Combining scale-out, line-by-line, and generation storage capabilities, DBMS enables us to solve complex architectural and technological problems facing the bank. It should be noted that ADB has the opportunity to further develop its functionality, including in the areas of working with sensitive data, which is one of the key guidelines for our bank - the reference for the defense industry industry - when choosing tools and designing information system architectures.}}

Arenadata DB can be integrated with third-party ETL tools, compatible with CASE data modeling tools (PowerDesigner), supports Big Data technologies. What was important for PSB, Arenadata provides its customers with technical support services in Russian, as well as a wide range of specialized training.

As of July 2024, PSB is implementing an updated data management architecture that meets the needs of the bank. Special attention is paid to the issues of data security, differentiation and control of access at the level of both role models, physical separation of data, and the use of various superimposed protections.

The project to build the Unified Data Warehouse of PJSC PSB is very difficult, but extremely interesting, which is being implemented in the era of change. The bank's team led by Ekaterina Varlamova demonstrates a decent level of professionalism even in such difficult conditions. told Pavel Ishkov, Director of Key Customers at Arenadata

The use of extended ECU data sets has already allowed the bank to solve analytical problems and achieve visible results. Thus, ad-hoc analysis based on ECD data is used by analysts to solve problems in the field of managing customer returns, attracting and retaining them, assessing bankruptcies, managing problem debts, as well as assessing the effectiveness of recovery.

Additional opportunities for accounting for business lines in analytical tasks appeared due to the joint use of extended arrays of source data of the bank's information systems, as well as arrays of data from external registers and aggregators-mass media within a single database (ECD). The use of ECD data allows the bank's specialists to quickly respond and eliminate deviations in the use of online platforms, mobile applications, and Internet banking.

The ECD data also developed a model for analyzing and identifying the risks of a state defense order (GOZ). GVA data, along with personal data, belong to classes of special sensitivity. The data security model developed within the framework of the ECD allows not only to carry out software monitoring of data work, but also to distribute working with them using developed "sensitivity" segments.

By the beginning of 2024, more than 200 analysts of various divisions of the TSOR GOZ bank, retail business, digital business, operations department, risks, marketing, debt collection, etc. are already working with the production ECD PSB. Access to data is regulated by the developed safety model and is determined by the role model according to the specified segments of "sensitivity." The model is developed and updated with the direct and constant participation of experts from the bank's information security department. To further optimize the effectiveness of decisions, the bank not only centralized the collection and accumulation of data, but also ensured the management of their quality. As of July 2024, the actual improvement in the quality of customer data of individuals by 56.1% has already been achieved.

Based on the results of the daily work of the system for monitoring the technical and business quality of ECD data, continuous work is underway to further optimize the quality of data in the bank's information systems. The ECU and Big Data PSB system is in the phase of active expansion of functionality.