Peter Reinbold - Project at a major Swiss bank
Peter Reinbold-Senior Consultant
Peter, could you please introduce yourself briefly?
I have been working in data management for 25 years, most of that time in the financial industry. I am currently working for in-factory in an agile team as an analyst, architect and technician on the development of a central data warehouse.
Which business does your project deal with?
In the long term, the data warehouse should map the entire business data pool of a Swiss financial services provider and form the basis for analytical reporting in all business areas. However, the architecture should also support operational processes. We started with compliance because this is where the greatest need currently exists.
Your project uses a DataVault architecture. Can you describe the approach and the main features?
Central models form the basis for development. They describe the data, structures and processes in technical and functional terms. The models form the basis for communication with the specialist departments, the central architecture and the development units. In this way, existing operational data can be successively brought into a business context in order to develop the system iteratively. And since most system components are generated from the model, it is always an up-to-date abstract representation of the implementation. The DataVault method supports this approach, as it only provides a small number of design patterns and therefore no complex individual developments are necessary.
Can you please briefly describe the reasons for this architecture and what advantages it has over traditional architecture?
The development of this system runs parallel to the transformation from the classic project model to agile methods. This requires a constant reaction to change. Data structures, technical or regulatory requirements, organization or processes are subject to constant change. The underlying architecture must be able to adapt to this and map changes correctly and support development accordingly.
How do you ensure that sponsors retain confidence in such a long-term project?
The interplay of DataVault architecture and agile methods enables the iterative development of manageable components and their commissioning. In this way, progress can be constantly presented and business value can be generated at a very early stage.
