POWER BI · SQL · SAP

Product Analytics & BI Foundation

The BI layer across the company's ERP landscape, covering the full order-to-delivery chain. Turned monthly Excel rituals into live views that saved €500K+, most visibly in procurement.

Role

Data Scientist

Timeline

2023 to 2024 · the foundation everything else ran on

The challenge

Data sat across multiple ERP systems and every department saw only its own slice. Procurement, product and operations decisions ran on partial pictures, and a product's lifecycle stage lived in seven different versions across seven functions.

The decision

Chose One shared semantic layer over Per-team dashboards

Ten dashboards with ten definitions of ‘revenue’ create arguments, not decisions. One model everyone queries creates a single version of the truth.

How I approached it

  • 01Mapped the full order-to-delivery chain across the ERP landscape before building anything: where data was born, where it broke, who decided what with it.
  • 02Built the shared data model and the Power BI layer on top, one set of definitions serving product management, order management, operations and procurement.
  • 03Shipped the product lifecycle report: seven functions, from PM to controlling, seeing the same stage for every product for the first time.
  • 04Turned reporting from a monthly ritual into live views, so decisions stopped waiting for the file to arrive.

Impact

  • €500K+ saved through data-driven decisions, most visibly in procurement.
  • Seven functions aligned on one product lifecycle view.
  • Decision delays and version-confusion errors cut across the order-to-delivery chain.
  • The data foundation the AI portfolio was later built on.

Stack

  • Power BI
  • SQL Server
  • SAP
  • Salesforce