Customer · Insurance
Nationale-Nederlanden rebuilt its non-life warehouse on Databricks with federated domains
The challenge: a legacy platform holding back change
For many years, NN relied on a large Oracle-based data warehouse that had steadily grown in complexity. As the platform expanded, it became increasingly difficult to operate and even harder to adapt. Operational costs were high, agility was limited, and small business changes required significant effort to implement.
These constraints slowed access to data and delayed the delivery of new insights. Reporting processes became harder to maintain, and performance suffered as data volumes and analytical use cases increased. It became clear that the existing platform could no longer support NN’s ambition to scale analytics across the organization.
The SIDAP program was initiated to address these issues. The goal was to give business teams direct access to both raw and modeled data in a scalable, cloud-native environment, while decommissioning the legacy Oracle warehouse without disrupting financial reporting, data processing, or visualization.
The approach: federated architecture with model-driven automation
Rather than lifting and shifting the existing environment, NN chose to rethink its data platform architecture. SIDAP was rebuilt on Databricks using a federated model in which individual domains could produce and manage their own data products while still operating within a centrally governed platform.
To accelerate delivery and establish a consistent development approach, NN selected VaultSpeed as the data automation engine. VaultSpeed generates production-ready code directly from the data model, introducing a model-driven workflow that aligns architects, data modelers, engineers, and operations teams.
This approach replaced manual development with automated and standardized patterns. VaultSpeed managed templates ensured consistency across domains while still allowing teams to work independently. Performance optimization techniques, including the use of domain-specific point-in-time tables, were embedded into the delivery process from the start.
The impact: faster delivery, better performance, stronger ownership
The results of SIDAP were immediate and measurable. Data marts that previously required long delivery cycles can now be delivered in 2-3 sprints. Complex table loads that once took up to 24 hours now complete in under 10 minutes.
The way data is consumed across NN has fundamentally changed. Non-financial business teams have become the largest group of users on the Lakehouse platform and now build their own data marts independently, without relying on central IT for small changes. This has significantly reduced bottlenecks and increased delivery speed.
Standardization across the platform has improved as VaultSpeed templates replaced ad hoc developer logic. Because all domains follow the same unified modeling approach, data used for regulatory reporting, including Solvency II, and actuarial pricing is aligned by design. This alignment has greatly reduced reconciliation effort and increased confidence in reporting.
Today, SIDAP supports 42 data sources, 24 of which are active, and manages a Data Vault of approximately 1.7 terabytes.
The takeaway: scaling data products without losing control
SIDAP has reshaped how Nationale-Nederlanden works with data. Faster delivery cycles, improved performance, and consistent modeling have given both business and technical teams the confidence to build and evolve data products at speed. Business teams now operate with greater independence while remaining within a centrally governed framework.
VaultSpeed is a cornerstone of this transformation. By automating data product development and enforcing standardized patterns, VaultSpeed enables NN to scale its data platform without accumulating technical debt. As the platform continues to evolve, NN is well positioned to support new data domains and broader analytical initiatives while maintaining quality, alignment, and business ownership.