Drawing upon the abilities of our team, supported by many years of experience gained in the financial sector, we offer comprehensive Enterprise Data Warehouse (EDWH) class solution design and implementation services.
Traditional Data Warehouses and Big Data
In today’s era of exploding data sources and volumes, brought about by the realization of the need for and benefits of analysis of mass, fast-changing and non-structural data, as well as the growing availability of Big Data class tools, traditional data warehouses are no longer the talk of the town for the broad IT community. However, the fact is that they are now, and will be in the future, the necessary foundation for the implementation of Business Intelligence solutions supporting many of the core business processes of every business, e.g.:
In time, all business areas are going to make more frequent and more extensive use of Big Data technologies, however, without abandoning the established, solid and proven traditional data warehouse solutions, in the first instance for the purpose of their own (internal) structural data.
EDWH architecture model
EDWH class solutions designed by us are based on the time-tested, canonical, multilayered architecture model, comprising:
It should, however, be pointed out that advanced technologies, including in-memory processing and storage of data, reduce the number of thematic databases which need to be implemented by making the resources of an Analytical Data Repository or an Operational Data Store available directly, which is more efficient.
In building data warehouses, we make use of both commercial technologies (Pentaho Enterprise, Oracle, Microsoft) and Open Source solutions (Talend, PostgreSQL, Mule ESB). Depending on the specific needs, a data warehouse architecture is supplemented with our proprietary reference data management products (Metastudio DRM) and/or data quality improvement products (Data Quality Studio).
We always compose proposed data warehouse solutions using:
Depending on customer preferences, we may implement a data warehouse using the traditional PRINCE2 methodology or the so called SCRUM approach. Regardless of the implementation methodology selected, we always strive for:
● Awareness that the end of an implementation project is the beginning of the maintenance process.