Building a Credit Data Warehouse for the Croatian Credit Bureau
Data Warehouse
Finance
- Data Warehouse Development and Implementation
- Credit Data Model Design
- Interbank Credit Data Exchange
- Bank-Specific Metrics Integration
- Confidential Data Access Interface
Unified credit data model
divergent credit data sources brought together
Interbank data exchange
banks access tailored customer credit data
Confidentiality maintained
data protected across the ecosystem
The Croatian Credit Bureau (HROK) needed a data warehouse developed and implemented to improve interbank data exchange across Croatian banking. Koios built a data framework capable of bringing together divergent credit data sources, along with an interface through which individual banks could access tailored customer credit data while confidentiality was maintained across the ecosystem.
The purpose of the system was to increase the accuracy of risk assessments, combat fraudulent activity, and promote prudent lending practices.
Croatian banking lacked the extensive credit data needed for well-informed lending decisions. Central to the problem was the absence of comprehensive credit histories — a gap that frequently resulted in flawed risk evaluations, because a bank assessing an applicant simply could not see the full picture of that customer’s credit behaviour elsewhere in the market.
Left unaddressed, that deficiency had the potential to expose financial institutions to heightened financial vulnerabilities, which is what underscored the need for a more robust system.
The approach centred on crafting a data framework that could blend divergent credit data sources into a coherent whole. Alongside it, Koios formulated an interface that would allow individual banks to access tailored customer credit data while upholding data confidentiality across the ecosystem — the condition on which any interbank exchange of credit data depends.
The work progressed through distinct phases. It began with the aggregation of the relevant data, moved on to refining the data model itself, and then enriched the assimilation process by appending proprietary bank-specific metrics, before culminating in operational implementation.
Tool selection was central to delivery. IBM DataStage managed data synchronisation and Oracle PL/SQL handled data processing, while data storage drew on the capabilities of PostgreSQL and Oracle databases. Aligning the build with the existing technical infrastructure meant those storage capabilities could be strengthened rather than replaced. SQL Server and Excel were also used in the course of the project.
- A data model built for precise risk assessment The established data model enhances how credit data is used for risk assessment and loan management.
- Flexibility for advanced creditworthiness algorithms Flexibility was built into the model so it can accommodate advanced operational algorithms for evaluating creditworthiness accurately.
- Improved loan operations and risk management The solution has proved consistently effective in improving loan operations and risk management strategies for the banks using it.
- Data scarcity overcome The challenges of data scarcity were addressed, leading to increased operational stability and strengthening overall financial robustness.
We'd love to hear about your analytics and digital optimisation goals
"*" indicates required fields
