Process mining and real-time BI in medical technology

Real-time analysis of complex process data from the operational database

The control and optimization of business processes requires a fast, error-free and comprehensive database. The challenges were complex: the process data was stored in an operational database that was not optimized for either complex or analytical queries in real time. Analytical queries on this database would have significantly impacted the performance of ongoing business operations. In addition, there was no easy way to centrally integrate different data sources, such as future machine data, and the existing data structure was not ideal for analytical activities and business intelligence (BI).

The solution: Real-time data warehouse

RISC Software GmbH developed a holistic solution based on modern data architecture concepts. Apache Kafka was chosen as the central component for real-time capable data transfer. This decouples the system speeds and enables the future integration of new sources. Debezium ensures that changes in the source database are continuously transferred to Kafka.

An analytically optimized data model was created for the data warehouse. Highly correlated tables were merged to speed up queries. The mapping to the target data model is carried out via a configuration file, which allows adjustments to be made without code changes.

Better solutions for customers

The data warehouse offers CADS GmbH decisive advantages for analysis and control. Thanks to the connection to Microsoft PowerBI, process data can be tracked in real time on dashboards, allowing direct intervention in the event of bottlenecks. The optimized data model significantly accelerates complex analytical queries. Above all, the data warehouse forms the high-performance basis for process mining, allowing internal processes to be analyzed and optimized in detail. The future-proof architecture is also designed for scaling and the simple integration of new data sources.