Abstract

Sotera is an Insurtech startup that was founded at Cambridge. After completing the Accelerate Cambridge program it joined Lloyds Lab at the Lloyds of London insurance market. Sotera started out looking at how to use Computer Vision to identify potentially looted antiquities. The business case pointed to insurance, and the company is now developing a tool that will be used in insurance and then offered free to the heritage sector to record at-risk objects quickly (eg. museums and collections at imminent risk from conflict). Sotera uses computer vision to identify high-value objects in a collection being insured. These objects are then analysed by a risk discriminator, which compiles the knowledge of experts like curators, insurers, law enforcement into an algorithm that calculates the risk of each object individually, and in context. This produces a risk report and risk pricing for complex policies of objects based on a granular understanding of the risk posed by each object within that specific context. The tool will change how insurers analyse risk within complex policies, such as museums, and for home contents policies of people with a collection of higher-value objects. Some questions we’re exploring include: - How can we infer size of objects from images? - How to deduce quantitative information from qualitative questions when interviewing experts about risk. - Simulation design: is there are faster way? - What is the best way to handle missing data: using averages or using random values from a distribution?

The talk will be lead by the Sotera team: Dr Trevor Maynard, currently Executive Director, Systemic Risks at the Cambridge Centre for Risk Studies, and former Head of Exposure Management and Reinsurance and Head of Innovation at Lloyds of London. Trevor is an actuary and mathematician with a PhD in Statistics from LSE ; Liz Marston, entrepreneur and Egyptologist who founded Sotera whilst at Cambridge; and Dr Tobias Stone, tech entrepreneur and former archaeologist with a PhD in Innovation