At Porsche we built computer models for complex events, like crashing a vehicle and understanding how hundreds of connected pieces of that vehicle would behave. We built large numerical models with hundreds and thousands of equations. We employed optimization and other engineering approaches to forecast on the computer how the structures would behave in real life. You could save a lot of money that way. It might cost $200,000 to build a model, but building a physical prototype costs $2 million. When Thomas and I started modelling GLI in 1998 we successfully employed a number of those engineering tools to simulate and back test strategies.
Later, I performed some large scale analysis for time and motion strategies for A.T. Kearney, seeking to gain efficiencies for a large insurer. We had millions of transactions, needed accuracy as well as speed, to correct errors and make sure we saved processing cost while the outcome was reliable and stable.
For a group of 30+ CEO’s at the World Economic Forum, I ran a 2-year study to investigate risk management and its impact on shareholder value in the global Engineering and Construction industry. These players build really big infrastructure from off-shore platforms to tunnels to skyscrapers. Our study directly led to the foundation of the Engineering and Construction Risk Institute. We developed comprehensive underpinnings to allow that industry to manage their risks, which are manifold, complex and with some cross-correlation, by quantifying and linking them more rigorously and explicitly than had been done prior.