The purpose of the Data WareHouse (DWH) project was to build a system, whose role would be to gather - in the form of time lines - and process analytically consistent data present in transaction systems of Všeobecná úverová banka (VUB). They are, first of all, data essential for the support of management and decision-making processes. The DWH system simultaneously allows to integrate information even if it is not integrated within individual VUB's production systems. The system was introduced into operation in 1999 and more than 100 staff members are currently working with it.

Data WareHouse is a storage area, separated from production systems, which contains transformed data from applications working in the framework of external sources. The transformation, however, is not without an added value, as its main role is to check the correctness of data as well as their consolidation in terms of the methodology and measurement units.

The Data Warehouse system is built on the Microsoft Windows NT platform, while the data are stored on Microsoft SQL Server 7.0.

The system contains:

1. an area of the bank's transaction objects, including:
1. Deposit and loan accounts
2. History of balances on them
3. History of transactions and the history of changes of accounts' properties
4. Area of the system of payment (domestic - foreign, documentary - non-documentary)
2. an area of information about VUB, including:
1. Relations between clients and their accounts
2. Information about references between banks and clients

Data Warehouse thus creates the possibility to get an overview of the complete portfolio of the client's transactions an services.

A highly important activity, done regularly, is the consolidation of data about clients. Because, for example, duplicate recording of clients takes place in individual (mutually not integrated) bank systems, in DWH there are a number of records related to a single client. In order to get a complex picture about the client's activities, it is necessary to link duplicate records about the same client and to formulate queries that take this into account (which, however, appears transparent to the end user). A part of this process is done automatically (search for candidates for consolidation according to own algorithm). However, the final finding out (or consolidation/non-consolidation) whether interlinking is actually possible has to be done manually.

The Data Warehouse system serves as a database for other Data Marts, which are designed for certain, narrowly specified areas, and it allows to carry out data mining. By concentrating on the data flow monitoring, with the DWH system it is possible to generate warning messages, distributed to key persons, e.g. by e-mail, which can inform about critical events more urgently than regular, e.g. daily reports about selected data.

Back to previous page.

Printer View | Tell a Friend | Contact us