Polly Mitchell-Guthrie
Polly Mitchell-Guthrie
Kinaxis
Track: Logistics
Making AI-Powered Forecasts Useful to Supply Chains
Artificial intelligence has tremendous promise, but it is not enough to sprinkle AI fairy dust on a forecast to make it useful for supply chains. After all, as statistician George Box famously told us, “All models are wrong, some are useful.” What does it take to make useful forecasting models, and what is the role that AI plays in doing so? How do we manage the balance of automation and human insight? How can we mitigate ‘wrong’ forecasts that occur from information failures before and after the model is built? Intelligent agility can mitigate tradeoffs and help manage disruptions to ensure that forecasting feeds into a process that delivers the value the C-suite expects these days from supply chain. This talk will outline five principles for success and provide practical examples for doing so.