Data Hatch
An LLM-powered application and renewal platform for a Lloyd’s coverholder. AI triage routes standard applications straight to issue, while a data-driven doctor risk model feeds price optimisation.
Dashed figures are placeholders pending confirmed engagement data.
Every application and renewal crossed an underwriter’s desk. Standard risks queued behind complex ones, quotes took days, and doctor risk profiling, a core line, depended on individual judgement that varied underwriter to underwriter.
As a Lloyd’s coverholder, there were no shortcuts available: the platform had to be faster while staying inside delegated-authority rules, with every decision defensible.
An LLM layer parses applications and renewals into structured records: fields, attachments, disclosures.
Professional registration, sanctions, and claims history are pulled automatically and attached as evidence.
A data-driven model scores each professional’s risk profile. Every score is explainable.
Risk scores feed directly into price optimisation, replacing flat rating with risk-reflective premiums.
Standard risks bind automatically with the right Lloyd’s wording. Elevated risks route to a senior underwriter with the profile and draft terms attached.
Renewals run through the same pipeline, and every decision is logged to Lloyd’s delegated-authority standards.
Underwriters only see risks that need judgement.
Triage, risk model, price optimisation, and the platform itself.
From inbox to issued policy on standard risks, zero manual review.
The triage layer replaces the review bottleneck.
Delegated-authority rules enforced inside the workflow.
Every premium carries a rationale.
Applications and renewals through one pipeline.
Dashed figures are placeholders pending confirmed engagement data.