Reinsurance Claims.
From notification to ceded settlement, automated.
Leah ingests treaties and claim notifications, applies treaty terms, computes ceded losses, surfaces interpretation issues, drafts settlement responses, and maintains the audit trail. Built for cedents and reinsurers.
Reinsurance claims are document-heavy. Manual triage is the cost center.
Treaty wordings vary across slips and renewals
Every renewal cycle introduces new wordings, manuscript clauses, and slip variations. Coverage triggers, exclusions, and aggregation language drift between layers and years. Claims teams reread the same treaty types repeatedly to confirm what actually applies.
Loss notifications buried in email and bordereaux
Cedent notifications arrive as PDFs, scanned forms, broker emails, and quarterly bordereaux. By the time a claim file is opened, the underlying notice has been forwarded across three inboxes and matched to the wrong treaty layer twice.
Retention layer math is manual
Working out which layer a loss falls into, where the retention sits, and how the loss participates across quota share, surplus, and excess of loss treaties is done in spreadsheets. Errors compound as more layers and years are involved.
Follow-the-fortunes interpretation is inconsistent
Whether a cedent settlement binds the reinsurer depends on treaty language, jurisdiction, and the reasonableness of the underlying decision. The same fact pattern is treated differently by different reviewers because the framework lives in people, not in process.
Settlement responses drafted from zero
Each ceded settlement letter is rebuilt from scratch, with treaty references retyped, layer math rechecked, and reservation language recomposed. The institutional drafting library exists in three colleagues' inboxes.
Audit trail fragmented across systems
Treaty documents live in one repository, claim notifications in another, ceded calculations in spreadsheets, and decisions in email threads. When auditors or counterparties ask how a ceded loss was determined, reconstruction takes weeks.
From slip to structured treaty data
Leah ingests every treaty document you have, slips, wordings, schedules, endorsements, and structures the terms that drive claims handling. Coverage triggers, retentions, limits, layer participation, exclusions, aggregation language, and follow clauses become queryable data, with versioning across renewals.
“Leah turned our treaty file room into a structured library. Layer math that used to take an afternoon is now a query against the program.”
Head of Claims, Global Reinsurer
Five steps from cedent notification to ceded settlement
Leah integrates with the systems your claims team already uses. No rip and replace. Value from the first notification ingested.
Connect
Leah integrates with your treaty repository, claim system, broker portals, and email. Treaty wordings, slips, claim files, and bordereaux flow into a single intelligence layer without replacing existing systems.
Extract Treaty Terms
Every treaty is read and structured. Attachment points, retentions, limits, follow clauses, aggregation language, and exclusions become queryable data, versioned across renewals.
Triage Loss Notifications
Notifications from email, bordereaux, and broker portals are ingested, matched to the right program and layer, and opened as claim files with full context. Late notices and duplicates are flagged.
Compute Ceded Losses
Each loss is run through the treaty tower. Quota share, surplus, and excess of loss layers are applied in order. Aggregation and reinstatements are tracked. Ceded numbers are produced with reasoning shown.
Draft Settlement Response
Leah drafts the settlement letter with treaty citations, layer math, and follow-the-fortunes analysis. Reviewers edit. Every step is logged for the audit trail.
Got Questions? Get Answers.
Leah reads the actual treaty document, not a structured template. Manuscript clauses, bespoke aggregation language, and syndicate-specific wording are extracted with citations to the source. Where extraction confidence is below threshold, the term is flagged for human review before it enters the layer engine.
No. Leah operates on top of your existing claim system, treaty repository, and broker portal feeds. Documents and records continue to live where they live today. Leah reads, structures, computes ceded losses, and writes back triage decisions, drafts, and audit entries. There is no rip and replace.
Each renewal year is tied to the underlying program in Leah. When a loss notification arrives, the date of loss, cedent, and line of business are used to apply the version of the treaty that was on risk at the time. Mid-term endorsements are honored where they affect the loss in question.
Leah applies a structured framework drawn from the specific treaty wording in force, supplemented with jurisdiction and case-law context where relevant. Each cedent settlement is reviewed for reasonableness against the underlying claim facts. Borderline cases are flagged with a reasoned recommendation rather than auto-cleared.
Leah surfaces the ambiguity with citations to the conflicting wording and a recommended interpretation. The decision stays with the reviewer or counsel. The recommendation, the override, and the rationale are all recorded so similar cases are handled consistently going forward.
Yes. Leah is deployed by global reinsurers, composite insurers, and Lloyd's syndicates with strict data security requirements. Treaty content does not train Leah's underlying models. Customer data is encrypted in transit and at rest. SOC 2 Type II, GDPR, and ISO 27001 aligned. Private instance deployment is available for customers with strict data isolation requirements.
