Consolidating production quality reporting on Databricks and Power BI

Consolidating BI on a modern foundation

Mirka’s business intelligence had grown across multiple platforms over the years – with Qlik anchoring their production test reporting. As new product lines, test systems, and data formats were added, the analytics landscape kept expanding.

To support future analytical needs and improve operational efficiency, Mirka decided to standardize its BI platform around Power BI and Databricks.

The new foundation had to preserve the meaning of the existing Qlik solution while improving scalability, governance, and usability.

Proekspert’s role: building Mirka’s Power BI-ready data foundation

Proekspert helped translate the Qlik-based reporting logic into a Databricks and Power BI architecture built around dimensional modeling best practices.

How we delivered:

• Analyzed the existing Qlik scripts, QVD outputs, and source data structures.
• Designed a Databricks medallion architecture from raw data to Power BI-ready gold tables.
• Unified production data feeds into a consistent analytical structure.
• Built a Kimball-style star schema with conformed dimensions and fact tables.
• Prepared the model for Power BI using clear relationships, reusable measures, and incremental refresh-friendly design.
• Tuned the Power BI model and reports following platform best practices.

Delivered features and impact:

• A single source of truth for production test measurements across product domains.
• Standardized handling of test parameters, limits, anomalies, products, orders, suppliers, serial numbers, and data sources.
• A simpler Power BI semantic model with one unified fact table instead of several disconnected reporting structures.
• A scalable Databricks foundation that supports future data sources and reporting needs.

Notably, the resulting Power BI model handles more than 500 million rows of production test data – governed, performant, and ready to extend.

A governed foundation for production quality analytics

Mirka’s business-critical test analytics now run on a modern, governed Databricks and Power BI platform – replacing the Qlik scripts and QVD files that had powered the previous setup.

Production test data from multiple domains flows into one consistent reporting model. Power BI users can analyze quality trends across products, test rigs, test parameters, and time – with a shared model keeping the numbers consistent across reports.

The Databricks foundation underneath runs faster, is easier to govern, and absorbs new data sources without architectural rework. For the BI team, that means reports that are quicker to build and extend, with fewer hidden dependencies to chase down later.

Overall, Mirka now has a maintainable, modern foundation that grows with the business and is well positioned to support future analytical needs.

Impact for the client’s business

Migration gives Mirka a future-ready analytics foundation that supports faster quality decisions, lower operating costs, and the capacity to grow without another rebuild.

• Faster, more confident production quality decisions. Unified test data across products, lines, and test parameters helps the quality team catch anomalies sooner and ship with greater confidence – the foundation Mirka’s reputation rests on.  
• A leaner cost base for analytics. Consolidating onto Power BI streamlines the analytics environment and enables more efficient use of development resources while providing a scalable foundation for future growth. 
• A platform ready for the next decade. Databricks and Power BI scale to new product lines, test systems, and analytics use cases (including AI), without forcing another rebuild as the business grows. 

Client

Mirka is a global family-owned company. They are dedicated to the finish, always seeking to renew and discover innovative solutions in surface treatment technology. With decades of experience, they are a world leader in surface finishing technology, offering innovative solutions for surface finishing and precision grinding.

Technologies

Databricks, SQL, Power BI, Qlik, Kimball star schema, medallion architecture