A recruitment consultant will contact you soon.

Air France – decision-making IS associated with a data warehouse

Air France's Data Warehouse for interactions between the different business lines and the sharing of consistent data

Context & challenges

A company like Air France expects its management tool to take on board interactions between its various fields of expertise and share coherent data gathered from the various activities carried out within the company.

Air France has chosen to pool its information. This choice makes it possible to provide a decision-making support tool consistent with the company’s various responsibilities, delivering a single version of truth at any particular moment.

This principle of sharing must be integrated into the decision-making information system built around a data warehouse accessible to the entire company: the Air France company data warehouse.

Average annual volume is 3,000 billion for the DWH team, including 1,000 billion for Traffic.

The aim: to monitor the activity of Air France’s ten main functional areas (commercial, operations, HR, Industry, etc.).


The stages of the project have been the analysis of existing data and professional needs as well as creating architecture/technical design files and, as required, the creation of a POC so as to select the technologies and/or offerings capable of meeting needs. The next stage is then to offer solutions enabling further progress to be made in data usage (Extraction, Transformation and Loading).

The final solution chosen consists of loading all Air France Traffic data into Teradata tables so that dashboards can be produced using Business Objects reporting tools.

The ETL PowerCenter extracts data from large flat files supplied by “operational applications” and then transforms it (format integrity check, etc.) ready for loading.

Transformation and integration into more complex DBMS data tables are carried out using FastLoad&BteqTeradata rather than ETL, mainly due to performance issues.


An ability to provide end users with accessible, coherent and “easy-to-use” data despite the high volumes involved and their considerable functional complexity.

A service including production monitoring of data integration lines, possibilities of carrying out corrective and evolving procedures, the ability to update and optimise databases, as well as provide support for end users.


Le point de vue de l'expert


Method and resources

Operation length and number of days: 

Infotel maintenance for the last 6 months
8 ETPs carrying out the BI maintenance

Expected deliverables:

Maintenance and improvement of Traffic data warehouse information integration lines.


Technical environments

Systems: UNIX (SunOs 5.8), Windows 2000 or XP

Database: TERADATA v2r6.2

Data integration: InformaticaPowerCenter v7.1.3
FastLoad, Bteq and FastExport TERADATA

Business intelligence: Business Objects Platform v5.1.8
Business Objects platform XIr2sp4