Datawarehouse: fast and flexible insight into your business processes

Background

Developing a sustainable business intelligence (BI) solution for a client in Rotterdam; a big project we have been working on recently. We designed a structured enterprise data warehouse to store all business data. The visualisations and analyses were then worked out in Microsoft Power BI. A solution with which the client made a big leap in business intelligence maturity.

2O years of data

With this solution, the client gained a quick and accurate insight into daily business processes, as the required data no longer had to be extracted from laborious reports. In addition, with the implementation of the data warehouse, the client has gained access to a treasure trove of almost 20 years of data, which allows easy and ad hoc analysis of long-term performance.

This is how we did it

Using our standard operating procedure and our knowledge of business processes, we first designed a data vault modelled structure for storing all relevant raw data. A data vault model can integrate data from multiple sources, making it ideally suited as the basis for an enterprise data warehouse. Data from the data vault model is then transformed into star models consisting of dimensions and facts. Microsoft Power BI ultimately uses the star models to visualise and analyse the data.

Keep it simple

Our vision for business intelligence? Keep it simple. Handle clear steps in the process from source data to intelligence. We regularly come across situations where clients have started to access source systems themselves using, for example, Microsoft Power BI or Tableau. In itself, there is nothing wrong with this, but it often results in a kind of copy of the source system, because there is no expertise in-house to create really suitable data models for business intelligence. In the longer term, the often complex data model of the source systems (and therefore the complex copy model) gets in the way of managing and developing the dashboards. By using discrete, clear steps in the process from source data to intelligence, you create a future-proof business intelligence solution that dashboard developers and data scientists can work well with.