Statefleet – NSW Government (Demand Forecasting) Oct 05 – Nov 05

Subcontracted by SAS Australia as the sole consultant for the analytical component of a large infrastructure project. Previous attempts to model used car prices had proved unsuccessful and the predictions of commercial operators Glasses Guide and Red Book lacked the accuracy required by StateFleet.

Novel data manipulation combined with data mining techniques were used to forecast residual value of state owned and leased motor vehicles at disposal prior to lease commencing. Created automated forecasting tool programmed in SAS which was deployed as part of the infrastructure project. This methodology and forecasting tool has been proposed to other State Governments and agencies by SAS.

Vodafone

>Developed a range of propensity models, including a major churn model project, that provided marked improvements over the incumbent/current methods. Remodeled the techniques and procedures for simplifying the modeling process within Vodafone. This included automating as much as possible, particularly the data collection and transformation.