Claims Root Cause of Loss
The Challenge
This project aims to unlock the value of historical, unstructured property insurance claim data, such as adjuster notes, estimates, and engineering reports, by converting it into structured, actionable insights. Using a retrieval-augmented generation (RAG) solution, the initiative will focus on two core use cases. First, it will enable detailed cause of loss analysis, helping insurers identify systemic patterns of damage, scale preventive programs for policyholders, reduce claims payouts, and contribute to broader improvements in construction practices and building code standards. Second, it will support automated recovery referrals by identifying liability linked to defective products or negligent contractors, improving recovery outcomes and promoting accountability. Together, these innovations aim to reduce environmental impact, lower financial exposure, and modernize how the industry prevents and manages property damage.

“This project allows us to revisit thousands of past property claims not as isolated events, but as a strategic knowledge base. By structuring this data, we aim to uncover root causes, detect systemic failures, and support more informed decisions, both in preventing future losses and pursuing meaningful recovery.”
– Michelle Laidlaw, AVP National Product Portfolio
Investment
$
0.9
M
Scale AI investment
$
2.3
M
Total investment
Partners



