Risk Flow Digest

Welcome to the second edition of Cytora’s Risk Flow Digest, where we share best practices on how insurers can digitize and optimize their risk processing workflows.

Large Language Models are getting global backing

Financial Times predicts that the global market for generative AI will grow from $6bn this year to $59bn in 2028, new forecasts show, as the technology is deployed in industries ranging from marketing to insurance. Generative AI is also finding a market in legal services and insurance by helping to automate laborious but important tasks — such as gathering, analyzing and creating documentation. This can include drafting large commercial contracts and spotting risks, such as irregularities in documents.?

Richard Hartley: “Insurance companies are bombarded with??information and all of it is different. Generative AI, at extremely low cost, can read all of this, digitise it, standardise it and highlight areas that are unusual.”?

Full article available here.

Making Risk Flow: Navigating the Constantly Evolving Cyber Insurance Market?

Check out the latest episode of our Making Risk Flow podcast on Navigating the Constantly Evolving Cyber Insurance Market, hosted by our very own Juan de Castro and featuring Meghan Hannes, Head of US Underwriting Management, Cyber Risks at Beazley. During the course of the discussion, Juan and Meghan cover the dynamic cyber insurance market, the evolving role of underwriters, and the role of risk controls and loss behavior in cyber insurance. Listen to the full episode here.

Praedicat and Veridion joined the Cytora data ecosystem

Cytora is welcoming two fantastic partners to join our ecosystem of data and insight providers: Praedicat, the world’s leading casualty emerging risk analytics and modeling company and Veridion a provider of AI-powered business data and market intelligence.?

Praedicat's and Veridion’s cutting-edge data, analytics, and models will be pre-integrated into the Cytora platform enabling customers to access and operationalise data sources directly into their risk processing flows. Read the full press release on Praedicat and Veridion.

New digitization features create further capacity and control for risk professionals?

• Section Detection enables risk professionals to easily navigate between all field candidates and documents that form part of an insurance transaction, such as a submission. Our new LLM powered “sections detection” feature automatically processes attachments and splits them into risk sections. This enables underwriters to achieve immediate context between extracted fields and the original submission and coupled with semantic search allows underwriters to perform rapid, cross-document queries to maximize their risk analysis time and minimize any time-consuming, manual processes.
• Approval workflows empower subject matter experts to review and approve training examples intended for model usage. This grants control to subject matter experts (including underwriting managers who have deep understanding of specific lines of business) over model performance, especially on complex risk fields which are semantically different across risk submissions. Approval workflows dynamically present training examples for review and approval, including highlighting instances where fields are similar yet not identical to the target schema fields.