Implementation of CS::BM Corporate Data Warehouse in OTP Bank

CHALLENGE

One of the priority tasks in OTP Bank’s IT strategy was replacing the existing data warehouse with an industrial solution. CS company was selected as the supplier, with whom the bank has been cooperating since 2009.

Among the implementation tasks were:

• Implement a reliable modern corporate data warehouse as a single source of truth in the bank: with a subject-oriented data warehouse model covering the maximum possible number of entities, and data quality control.
• Migrate the functionality and data from two existing CDWs to a single modernized CDW. Transfer existing data marts from the bank’s data warehouse.
• Implement data loading on T+1 / (T = transaction date).
• Increase the amount of data/entities needed by business users in the warehouse.
• Obtain a fault-tolerant managed solution.

SOLUTION

To address the tasks, CS company developed:

• Data loading management mechanisms;
• Data quality control module;
• Solutions for providing data for group reporting.

The following was also completed:

• Enhancements for controlling the data loading process and data mart calculations;
• Refactoring of the bank’s previously used data warehouse code.

The CS::BM solution is built on the traditional methodology for large DWHs proposed by Bill Inmon. The methodology is based on a centralized repository of all corporate data, which then serves as the foundation for data marts.
The DBMS used is Oracle 12C.
The ETL/ELT tool is Oracle Data Integrator (ODI) 12C.
The data mart visualization tool is Oracle Business Intelligence (BI) 12C.

Sergiy Rachkov, Business Analyst at CS company:

The DWH in OTP Bank is populated from more than a dozen sources. Each system has its own data loading regulations. Analysis of the relationships between loads is performed at the data warehouse core level to ensure the correct approach to data mart implementation.
On a daily basis, data is extracted from about 700 tables, and more than 50 different data marts are formed, which are used both directly by users through Oracle BI and for integration with data-receiving systems. About eight systems consume data calculated in the data warehouse. These systems are built on different databases and use different approaches to data loading. All this necessitated the development of completely different inter-system interfaces that fit into the overall solution architecture.

Oleh Rozum, Head of DWH Development, Project Manager, OTP Bank:

The Bank’s project team and CS company faced ambitious tasks:

• consolidate analytical reporting at the level of a single DWH solution using an optimal amount of resources;
• update the hardware and software components of the DWH solution;
• migrate and modernize data marts from previous DWH solutions, as well as transfer labor-intensive analytical reporting from Master systems to the new DWH;
• document the new DWH solution;
• eliminate architectural and logical “gaps” of previous solutions;
• ensure data storage security and safe data operations;
• ensure efficient data storage;
• implement a data quality control model;
• establish the foundation for the “Business Self Service” reporting approach based on the BI solution.

Thanks to joint efforts, cohesion, mutual understanding, and result-oriented work, the project was successfully completed.

The implementation used CS’s proprietary Knowledge Modules for Oracle Data Integrator:

• Loading Knowledge Modules (LKM).
  • For extracting from systems on Oracle DB.
  • For extracting from systems on MS-SQL.
  • For extracting files from the file system, ftp.
• Integration Knowledge Modules (IKM).
  • For loading to DWH.
  • For forming data marts.
  • For exporting prepared information to other information systems.

RESULTS

Компанія CS успішно завершила масштабний проект із впровадження корпоративного сховища даних CS::BM в OTP Bank.

В результаті виконаного проєкту:

• Банк отримав рішення з супроводом.
• Вітрини CS::BM розраховуються швидше порівняно з попереднім рішенням, що було критично для банку.
• Забезпечено своєчасне подання групової звітності.
• Сформовано оптимальний процес роботи з DWH — супровід, локальний розвиток, вендор.
• Можливість розширення — за необхідності до системи можна підключати різні нові джерела даних.
• Єдина компанія-виробник для ключових систем, які використовуються в ІТ-архітектурі, що дозволяє завчасно імпплементувати зміни до DWH за фактом змін у core-системах.
• Винос ряду ресурсоємних задач із систем-джерел до DWH.