The Lloyd’s of London insurance market, one of the longest-standing institutions in the insurance market, known for its centuries-old traditions and ability to insure unique and complex risks, is witnessing a wave of innovation. At the forefront of that wave is Ki, a digital-native syndicate that’s reshaping the landscape of specialty insurance through algorithmic underwriting.
This talk explores Ki’s journey from concept to a billion-dollar business in just three years. We’ll dive into the unique data science challenges of specialty insurance – from microscopic dataset sizes and rare claims events to unstructured data trapped in PDFs and handwritten notes. You’ll discover how we built algorithms that can assess complex risks and determine optimal line sizes in seconds rather than days, despite these constraints.
We’ll share our technical evolution across risk assessment modelling and portfolio optimisation, driven by our pragmatic “start simple and iterate” approach in a domain where traditional ML best practices often don’t apply. Along the way, we’ll reveal technical and cultural challenges of introducing Data Science to Specialty Insurance industry veterans.
Whether you’re interested in ML with limited data, sequential decision-making under uncertainty, or applying data science in traditional industries, this talk offers insights from the frontlines of InsurTech disruption.
Technical level: High Level/overview
Session Length: 40 minutes
Supported by
