​

The theory side of BI concerns itself with the logical "how" of getting data efficiently from company source systems, modelling the data for ease of access and the presentation of data to end-users that tells a coherent story that can aid the company's decision making process. 

 

Of these theories, two giants have emerged and have been embraced globally as the definitive standards for implementing BI architectures; the Kimball Methodology and the Inmon Method. 

​

Though different in approach, both methods focus on separating the data required for analysis from its native source and depositing the data in it's own dedicated centralised reporting environment, in which the resultant data is modeled into dimensional data marts that provide fast and easy access for data analysis.

​

This process is automated and is commonly known as a relational Data Warehouse. This forms the foundation layer of the BI Architecture. 

​

The data can then be digested by supercharged reporting databases known as OLAP and Tabular databases, which can be configured with pre-built calculations and deliver lightning fast response times. This makes up the middle layer of the architecture.

​

Finally you have reporting clients that connect to the foundation and middle layers, to present the data to the business users in the form of static or interactive reports, spreadsheets and data visualisation software. Making up the top layer of the architecture.

​

The sum of these layers make up the modern BI Solution architecture.

​

- Note, the beauty of this standard method is that they're technologically agnostic, meaning companies can take their pick of BI technology vendors and still implement a standardised and effective BI solution architecture.