I started life as a physicist, completing an undergrad and PhD in physics at Oxford. In my PhD, I focussed on both theoretical and experimental studies of superconductors and magnetic monopoles. After my PhD, I wanted to try something a little more applied. I am now a Senior Deep Learning Researcher at Tractable, applying cutting-edge computer vision algorithms in the auto insurance industry.
Imagine the scene: your car is damaged; your freedom is interrupted. You're faced with two paths: either have your insurance claim sent to a human, who may take weeks to make a settlement, or send your claim to an AI, which can settle in minutes. In this talk, I explore how the latter is possible today. Through the use of cutting-edge image recognition and segmentation techniques combined with expert logic, I will present how it is possible to emulate expert-level tasks in auto insurance using AI.