Data Mesh Enabled Data-driven Decision-making and Ensured Long-term Sustainability

Background and Challenge Story

In the ever-evolving landscape of the automotive industry, Adastra’s journey with a leading global automobile manufacturer was marked by a significant transformation. Faced with challenges typical to a large, heterogeneous system landscape; the client grappled with data silos, a lack of findability, and a provisioning process that often extended over many months. This prompted Adastra to step in as a transformative force that guided the client on Data strategy, Data Culture, Data Governance and Tech Stack to reshape the data landscape.

The success story unfolded within the automotive giant’s financial services division This division encompasses dealer and retail financing; leasing; direct banking; and insurance business. The system landscape was characterized by its heterogeneity which hindered efficient data management. Large data silos created an environment where crucial data was difficult to locate, which  led to delayed decision-making. The provisioning process, – essential for data utilization – was a cumbersome task that took months or even years to complete. The need for a paradigm shift was evident, and Adastra identified the opportunity to establish a Data Mesh as a lighthouse for service and quality improvement. The focus was on the vehicle life cycle and residual value analysis; leveraging ML (Machine Learning) models on historical sales; and leasing data and vehicle information.

Several challenges emerged during the initial stages:

• Data owners were hesitant to take responsibility for their data products
• Complications with self-service developmen
• Challenging  data shopping proces
• The landscape was characterized by legacy systems and a highly heterogeneous tech stack
• Complicated data lineage
• A clear separation between enterprise architecture management and data governance worsened the situation

Solution Story

Based on the information gathered during the discovery phase, it became clear that the strategic way forward was to implement a Data Mesh for addressing the above-mentioned challenges. The Data Mesh was organized into multiple data layers, utilizing the publish-subscribe method to connect data producers with data consumers. This architecture facilitated the creation of consumer-oriented data products (data marts); derived data products; and raw data products, that each served a distinct purpose within the organization.

Data Culture and Governance aspects

In navigating these challenges, Adastra recognized the pivotal role of fostering a data culture. Setting incentives to encourage data ownership – coupled with coaching, training, and proactive monitoring – played a crucial role. The introduction of a data product starter kit and templates streamlined self-service development. Immediate initiation of data governance, component reuse from existing systems, and harmonization of the tech stack were pivotal in overcoming legacy issues

Technological Guidance

Adastra implemented a solution aimed at supporting decision-making. ML-supported data analytics on historical dealer and contract data provided the basis for informed decision-making in sales, which further contributed to added value.

Benefits

Adastra’s partnership with the automotive giant not only addressed immediate challenges, but also laid the foundation for a data-driven culture. This new foundation led to enhanced decision-making and promoted long-term sustainability. The success achieved in this transformation serves as a testament to the power of embracing Data Mesh principles and establishing a robust data governance framework in navigating the complexities of a large-scale, heterogeneous system landscape.