Quote Faster. Underwrite Smarter. Grow Profitably.

The leading platform for reinsurers/insurers/MGAs who compete on speed, accuracy, and operational excellence.

Inbox connector
Microsoft 365
submissions@acme-re.com
  • placing@broker.example
    SOV · Acme Warehouse
  • london@broker.example
    Slip · Q3 renewal
  • claims@broker.example
    Loss run · 2019-2021
Connected · read-only · 1 folder
app.reinsured.ai/packs/submission-1842
87Ready
Pack readiness
Submission pack assembled
FieldValueConf.Source
Limit$25,000,000
94%
p.14
TIV$142,800,000
91%
Sheet: SOV!C12
OccupancyWarehouse
88%
p.3
Deductible$250,000
96%
p.14
Gap report
  • SOV occupancy uncoded (3 locations)
  • Loss run years 2019-2021 missing
  • Target premium not stated

OAuth · read-only scope · full audit log · revoke anytime

Plugs into where you already work

Microsoft 365
Gmail
SharePoint / OneDrive
ServiceNow
Document management
Policy admin
MCP: Model Context Protocol

Any internal system can become a source or a destination through open MCP connectors, with no rip-and-replace and no data exports.

Every connector is org-scoped, permissioned per team, and every read is written to the audit trail.

Every field shows its work.

Each extracted value carries its source document, page or cell, a confidence score, and any conflicting values from other documents. Your underwriters review evidence. They don't re-key data.

How source-grounded extraction works

Limit
structure.limit
94%
$25,000,000
Extraction certainty94%
Origin fileSubmission vF.xlsx
Sheet / cellStructure (C7)
Submission vF.xlsx / Sheet 'Structure' C7
Conflicting values detected
Slip p.3$20,000,000
conflicts with: Slip p.3 ($20M)

Specialised Agent Catalog

Composable AI agents designed for specific reinsurance/insurance workflows. Deploy individually or combine for end-to-end automation.

Plus 19 specialist agents across exposure, recovery, ILS and capital →

Underwriting Agent

Job: compare a traced pack to stated guidelines and list what is in, out, or missing. It does not silently bind a layer.

Contract Agent

Job: pull named insured, period, limits, deductibles, and endorsements from wording PDFs with a source span, or list the gap.

Schedule Agent

Job: turn an SOV into a summable location list with as-at date, or flag rows that cannot be totalled.

Actuarial Agent

Job: extract cited assumptions and figures from actuarial PDFs so pricing conversations use the same numbers as the report.

Triage Agent

Job: classify an inbound pack by class and completeness, then route it. Incomplete files stay on the chase list, not in a fake-complete queue.

Exposure Agent

Job: aggregate traced locations and limits into an exposure view, and list schedules that still have no as-at date.

Retention Agent

Job: assemble retention and cession inputs from cited wordings and bordereaux so a human can choose structure. It does not pick the programme.

Capital Agent

Job: gather filing-relevant fields that already exist in the pack and list what is still missing for a capital conversation.

Accumulation Agent

Job: roll traced locations into peril accumulations and surface schedules that cannot be geocoded or summed.

Retrocession Agent

Job: extract retrocession structure, attachment, and reporting fields from the contracts you already have, with gaps instead of guessed layers.

Quote Agent

Job: draft a quote memo from traced pack fields and list what still blocks a number. It does not invent a limit to look complete.

Claims Agent

Job: map loss notices and claims bordereaux to treaty terms, then list breaks, duplicates, and missing as-at dates.

Market Agent

Job: keep market, broker, and placement notes next to the pack so a desk is not pricing from a covering email.

Portfolio Agent

Job: show how traced in-force limits sit against stated appetite, and list policies whose schedules still do not sum.

Renewal Agent

Job: compare expiring wording and bordereaux year-cuts to the renewal pack, and list fields that changed without a span.

Performance Agent

Job: compute ratios from premium and claims files that already reconcile, and refuse a combined ratio when the year-cut is still a gap.

Clash Agent

Job: find where the same event language could hit more than one treaty in the pack, and cite the clauses rather than infer them.

Commutation Agent

Job: extract outstanding, paid, and commutation language from the files you have, then list missing reserve as-at dates.

Sidecar Agent

Job: pull sidecar reporting fields from agreements and bordereaux, and list collateral or performance items still without a source.

ILS Agent

Job: extract trigger, term, and reporting language from ILS documents, and leave unsourced return figures as gaps.

Architecture

The 4-Layer Insurance Intelligence Stack

Most AI tools in insurance operate at a single layer: extracting documents or automating a task. Reinsured.AI is architected across all four layers, with the Context Cloud as the connective intelligence that no point-solution vendor can replicate.

Explore the full architecture
01
System of Record
Policy, claims, contracts, accounting
Guidewire / Sapiens / Cenata / Duck Creek
Partner Systems
02
System of Context
Business brain across all systems
Reinsured.AI Context Cloud
Differentiator
03
System of Decisions
What should happen next?
Reinsured.AI AI Agents
Our Core
04
System of Actions
Execute: email, task, update system
ServiceNow + Reinsured.AI
Our Spoke
Towards AGI Frontier
Industry First

Organisational Sovereign AI for Insurance

Reinsure-8B is the world's first purpose-built small language model for the reinsurance industry, fine-tuned from Llama 3.1, trained on reinsurance workflows, treaty structures, and market language. Your organisation can run it fully sovereign, or consume it via a simple inference API.

Unlike generic LLMs applied to insurance, Reinsure-8B speaks your language out of the box. No prompt engineering, no hallucinated policy terms, no data leaving your perimeter.

Explore the model
Model
Reinsure-8B
by Reinsured.AI
HuggingFace
Base modelLlama 3.1 8B
Parameters8 Billion
DomainReinsurance / Insurance
TrainingFine-tuned on proprietary datasets
WeightsOpen on HuggingFace
SovereignInference API

Trust Architecture

Built for institutional reinsurance environments. Security, compliance, and data sovereignty are non-negotiable.

Data Sovereignty
Your treaty data never trains our models. Zero-knowledge architecture with client-side encryption. Full data residency control across jurisdictions.
Regulatory Coverage
Pre-configured for Lloyd's, Bermuda BMA, EU Solvency II, Swiss FINMA. Automated compliance reporting and audit trails built-in.
API-First Integration
RESTful and GraphQL APIs. Live in 48 hours with no legacy system replacement. Connect to existing bordereaux platforms, catastrophe models, and policy admin systems.
Security Certifications
SOC 2 Type II, ISO 27001, GDPR compliant. Penetration tested quarterly by independent third parties.
Market Intelligence
1,360

automated tests on every release

80

real-world document sets in our regression corpus

2,470

records benchmarked

100%

of extractions carry field-level provenance

Renewal season, measured in hours.

Quick Questions

Essential answers about AI automation in reinsurance operations.

What is Reinsured.AI?

Reinsured.AI provides specialized AI agents for global reinsurance operations. Our production-ready agents automate document processing, underwriting, pricing, compliance, and portfolio management, eliminating the 40% manual data entry waste that plagues reinsurance teams.

How quickly can we implement?

Standard integrations go live in 48 hours. Complex custom workflows may take 1-2 weeks. Unlike legacy system replacements, our API-first architecture layers on top of existing systems, so there's no rip-and-replace required.

What's the typical ROI timeline?

Most deployments show positive ROI within 3-6 months. Clients report 40% time savings on manual tasks, $7-9M annual operational savings, and 60% faster submission processing. The Bordereaux Agent alone saved one Zurich reinsurer $8.2M annually.

Is our data secure?

Yes. Your treaty data never trains our models. We're SOC 2 Type II and ISO 27001 certified with data encryption at rest and in transit. Support for data residency requirements (US, EU, APAC) and on-premise deployment for Lloyd's market participants.

Do we need to replace existing systems?

No. Our API-first architecture integrates via REST APIs and webhooks with existing policy admin, accounting, and document management systems. We layer on top of your current infrastructure: no rip-and-replace required.

Can agents handle both insurance and reinsurance?

Yes. While we specialize in reinsurance complexity (bordereaux, treaty structures, facultative slips), our agents also serve primary insurers and MGAs. We support property CAT, casualty, specialty lines, life & health, and ILS structures.