General Motors is the world's largest automaker and  has been the global industry sales leader for 75 years. GM employs about 327,000 people around the world and it manufactures its cars and trucks in 33 countries.  GM also operates GMAC Financial Services and it is one of the world's leading finance companies which offers automotive, residential and commercial financing and insurance. It's OnStar subsidiary is the industry leader in vehicle safety, security and information services.

• To build a rapid prototype for calculating true year-over-year savings on parts purchasing
  

• Data challenges included: • Parts merging into parent parts over the course of time, and Parts subdividing into multiple subcomponent parts over time
  

• Multi-currency and currency fluctuations over time
  

• Sheer volume of data for analysis
  

Ephibian analyzed the GM supplier/parts data structure. The parts data fluctuate over time, and the associated cost data fluctuate with changes in multi-national currency and exchange rates. The parts themselves can consolidate or subdivide over time; as the most simple example illustrates, a nut and bolt may at one time be labelled a single part, while months later –after an engineering or contract change --- they might be categorized as two distinct items.

We built a rapid prototype to calculate the true spend and actual savings accounting for fluctuations in parts nomenclature and variances in currency, over time. The rapid prototype was built with Microsoft Access and Visual Basic to operate against a subset of the data, and prove the concept.

Once demonstrated, a more scalable system was built in Perl to access GM’s full parts database resident within GM’s DB2 database.

The resulting implementation allowed the comparison of actual year-to-year performance matched against the financial numbers.

• More accurate analysis and reporting with full traceability resulting in millions of dollars saved annually