We helped de Volksbank by building a data warehouse to optimise their entire risk and financial reporting systems. Ultimately, an entire data control framework was created around the warehouse for data lineage, quality, correction and reconciliation.
- The Challenge: Develop data pipelines and data models with clear data lineage to accurately trace information from data source systems to financial and risk reports.
- The Approach: Align all parties involved and monitor the progress of the data sources at all stages of the project.
- The Solutions: Build a shared data warehouse for Finance and Risk that delivers data for the risk models and reports.
- Client Impact: Prove the entire data lineage, reduce the margin of conservatism, and develop a more efficient reporting process.
- Learnings: Recognise the importance of implementing a strong foundation of well-determined data.
The European Central Bank requires financial institutions to be able to trace all the information from data source systems to their financial and risk reports. However, the data lineage is not always clear or easily traceable.
Initially, de Volksbank tried to resolve this issue within their existing data warehouse, but ultimately, they decided that it is something they would have to build from scratch. Our team at ADC helped them optimise their entire data system. We did so by using new technologies to build a strong foundation with a clear data lineage feeding into the new data warehouse.
One of the main challenges that we faced was engaging with the risk management department. Prior to us joining the project, it had been continuously postponed leading to an understandable lack of engagement. However, their involvement was crucial, particularly in the testing phase to ensure that all aspects were in accordance with their standards and expectations.
To begin, we needed to align all parties involved, specifically the risk management and IT departments of de Volksbank. This allowed us to both meet and exceed their expectations by being transparent about our capabilities from the outset.
Additionally, we monitored the data sources and specific access layer, as well as the number of attributes that needed to be built. These were then monitored throughout the various stages of the project. By efficiently communicating these stages and addressing obstacles early, we gave the program director a secure grasp on the progress of building the data warehouse.
In summary, we were able to prove our trustworthiness and reliability by building relationships with sponsors within the organization. This resulted in increased engagement during the testing phase.
Through a combination of state-of-the-art technology and expert risk domain knowledge, we strategically helped de Volksbank with both complex choices in their blueprint and in-depth implementation. The ultimate deliverable was a data warehouse. It went into production and was able to deliver the data for the risk models.
Building an entirely new data warehouse was partially an unfamiliar concept for our client, resulting in many unknowns. However, we were able to find solutions through flexibility and creativity. Our team understood not only the business perspective of the risk management department, but also the technical perspective of the IT department who would ultimately need to build the data warehouse.
Earlier, attempts to build data warehouses had been made using, for example, bare SQL programming language. This is both a tedious and error prone method in which lineage is broken for complex pipelines. Instead, our team used PowerDesigner, a technology that writes SQL code itself from logical data modelling. In logical data modelling, lineage is preserved. This allowed for much more flexibility and fit within the client’s vision and design.
Impact for the Client
Ultimately, this project was crucial for protecting de Volksbank’s license to operate. Implementing an entirely new data warehouse with a data control framework around it allows them to prove the whole data lineage. Additionally, it shows the data that flows into the risk models that are used for reporting.
From a risk management perspective, removing the uncertainty around the data in the risk models reduces the margin of conservatism. As a result, it also reduces the capital the bank is obligated to hold. This is because the required capital is based on the internal models.
Finally, this project allowed our client to become more efficient internally, particularly with reporting. The reporting process is now more accurate, consistent, and timely within the bank. This allows for greater internal control over the entire process.
This project was a reminder of the importance of implementing a strong project foundation using data. It is necessary to have a structure in place where well-determined data, particularly through data gathering, lineage, and structuring can be used for things such as decision-making and developing models.
Curious what we can do for your organisation?
Would you like to know more about this case or what Amsterdam Data Collective can do for you? Get in touch with Govert van Koningsveld of Amsterdam Data Collective via firstname.lastname@example.org or check our contact page.