The client wished to estimate the pricing model of competitors in order to undercut them in bids. Using historical data of previous successful and unsuccessful bids I was able to create a linear regression model for each of the companies key competitors, insuring their ability to undercut them at future auctions.
The linear model was based on a price per unit, along with haulage costs. My model made use of geo-api's to accurately estimate such costs for historic purchases and therefore deduce competitors price per unit.
This project consisted of consolidating/cleaning inconsistent historic data as the company was on the lower-end of the data maturity curve.
A dashboard was created on an internal server that provides real time estimations of these bids as the dataset updates in order to provide the directors with real time insight and notice immediately should their competitors behavior change.