Conception and implementation of an Enterprise DataWarehouse


A large group consisting of subsidiaries in different countries grouped into divisions (by product category).

Awareness of the cost and redundancy of extractions from source systems and warehouse modeling for local, divisional and global requirements.

Desire to capitalize on technical-functional knowledge and reduce the time-to-market of decision-making projects by providing a centralized, standardized and documented data warehouse, gathering all the company’s data, all source systems (SAP, HR, CRM, etc.).

Willingness to offer the possibility of intraday analyses for operational piloting, in addition to the traditional analyses on D+1.

Target User Population

BI Architects and developers, and all business departments

Succcinct description of the implementation of this application and its added value

Creation of a data warehouse consolidating the data coming from the following sources

  • 9 SAP systems (purchasing, sales, production, finance, etc.)

o Optimized modeling for reporting complex objects:

– BOM (Bills of material)

– classifications

– Hierarchies (Finance, Customers)

  • 3 HR systems (pay, career management, training)
  •  CRM
  •  Planning system (sales and production planning)
  • Ticketing system (management of incidents and evolutions for the IS)

More than 500 tables made available, with a shared and documented nomenclature.
Incremental refreshing of transactional data every two hours.


  • Decrease in the skill level and time required to implement new datamarts.
  • Reduction of impacts related to source system versioning (one batch of extractions to be modified)
  • Provision of previously impossible analyses (short-term monitoring of the service rate, etc.)

Process and Sourcing

Developments carried out with Informatica PowerCenter, the warehouse being supported by IBM DB2 DBMS. Incremental extraction from SAP systems is based on SLT technology from the editor.