Insurance & Carriers

Carrier-grade data work,built for your domain

Actuarial models that must stay reproducible. Claims history that cannot drift. Reserve positions that have to reconcile to the penny. Generic migration tooling does not understand any of it, so it hands the hard parts back to you.

837 EDI
Claims automation in production
Read, validate, write, audit
IFRS 17
Compliance migrations delivered
With audit evidence attached
90 days
Carrier M&A integration
Against a regulatory deadline
Zero
Downtime cutover
Wave-based with rollback
The architecture

One objective in.Your whole estate, moving.

The governed layer sits at the centre of your operations. It reads every signal across Policy admin systems, Claims intake, Underwriting data and the rest of your estate, and closes every loop back to your business and regulatory goals.

Policy adminsystems
Claimsintake
Underwritingdata
Fraudsignals
Customerhistory
Regulatoryfeeds
Riskmodels
Live

Insurance Governed Data Layer

IntelliBooks · Platform agnostic · Your cloud · On-prem capable

Ingestion & parsing
EDI, flat file and API sources normalised, including scanned documents
Governed models
Policy, claim and exposure conformed with grain agreed and documented
Agent registry
Insurance-specific agents for underwriting, claims and compliance
Orchestration
Wave-based migration with parallel running and rollback per wave
Reproducibility
Model inputs versioned so prior valuations re-run to the same answer
Human in the loop
Reserve and reserving-convention conflicts escalate to actuarial
Snowflake · Databricks · Redshift · BigQuery · Synapse · Iceberg — alongside the policy and claims systems you already run
Any warehouse · Any cloud · Any orchestrator · Bring or build agents
Quotesissued
Underwritingdecisions
Claimsresolved
Fraudflagged
Renewalsretained
Compliancereports
Serviceimproved
What does your organisation actually need?

The right question changes the answer.

Most insurance teams have modernised in pockets. We start with why those pockets never joined up, and what it takes to run the whole estate on one governed layer.

Policy history spans multiple systems and several decades, including products long since closed to new business. How much of it could you reconcile on demand today?


Agents map the legacy schema against the target, separate live products from run-off, and convert history in waves — each one reconciled against source before the next begins.

Actuarial models depend on inputs that change over time. If last year's valuation had to be reproduced exactly, could it be?


Model inputs are versioned alongside the data, so any prior valuation re-runs and returns the same number. We treat that as the acceptance test, not a nice-to-have.

IFRS 17 requires a data structure most estates were never designed to produce, against a date that does not move. Where does yours stand?


The restructuring is agent-executed and the evidence is generated as it runs, so transformation logic is documented at the point of execution rather than reconstructed for auditors later.

Acquired books arrive with their own reserving conventions. Who decides which convention survives, and is that decision written down?


Books are sequenced by risk and reporting impact, and every convention conflict surfaces to actuarial as an explicit decision with both positions shown. Nothing is resolved silently.

From objective to outcome

This is how the work actually runs.

01

Your policy and claims systems stay.

Your policy administration platform, your claims system, your reinsurance platform, your actuarial tooling. Connect what you have. Nothing migrates off its system of record.

Zero rip-and-replace
02

Scope backwards from the date that cannot move.

Not a wish list. A deadline. "Report under IFRS 17 on time." "Integrate the acquired book in ninety days." "Come off the legacy PAS before support ends." We work back from the fixed date and say plainly which parts are genuinely deliverable.

Deadline-first scoping
03

The right agents activate, inside limits you set.

Policy conversion, claim validation, reserve reconciliation, reinsurance recovery matching, regulatory aggregation. Autonomy per process: assistive for reserving, delegated for policy conversion, autonomous for validation. Convention conflicts escalate to actuarial.

Staged autonomy, earned not assumed
04

Quotes issue. Claims resolve. Reserves reconcile.

Every action carries its query and identity, and prior valuations still reproduce to the same number. Explainable at the moment it happened, to your team, your auditor or your regulator.

Explainable at execution
Coverage

We map to how you already run.

Tell us which of these hurts most and we start there — not with a platform rollout.

Policy administration

Historical policy conversion with per-wave reconciliation against the legacy system.

Claims migration

Heterogeneous claims models mapped with reserve positions preserved exactly.

837 EDI processing

Claim files read, validated, enriched and written to the system of record with audit.

Underwriting data

Risk, exposure and pricing inputs conformed so decisions are explainable.

Actuarial reserving

Reserve triangles and model inputs versioned for reproducible valuations.

IFRS 17 / LDTI

Regulatory restructuring with transformation logic evidenced as it executes.

Solvency II

Submission data assembled from primary sources with provenance intact.

Reinsurance

Treaty and recovery data modelled so ceded positions reconcile.

Broker & agency

Distribution feeds normalised across differing file formats and cadences.

Fraud & SIU

Claim patterns surfaced with the supporting evidence attached to each case.

Carrier M&A

Two estates merged with reserving-convention conflicts escalated, not hidden.

Regulatory reporting

Returns produced from governed data rather than reconciled spreadsheets.

The question you are already asking

Your data. Your platform. Your call.

It runs in your cloud

Deployed inside your own account and VPC, or on-premises where policyholder data residency requires it. Nothing leaves your perimeter during the work.

Governed at every step

Every agent action logged with its query and identity. PII handling and masking aligned to the HIPAA and GDPR controls carriers are held to.

You keep what we build

Models, pipelines, lineage and agent definitions are yours. Nothing depends on a proprietary runtime you would have to keep paying for.

Reproducibility is the contract

Prior valuations must re-run to the same answer after the move. That is the acceptance test we hold ourselves to, not a nice-to-have.

IFRS 17Migrations delivered
LDTIMigrations delivered
Solvency IIReporting supported
HIPAACompatible
SOC 2Aligned
GDPRReady architecture
Works with your stack
SnowflakeDatabricksAWS RedshiftGoogle BigQueryAzure SynapseApache IcebergdbtApache AirflowApache SparkNeo4jTrino837 / X12 EDI

If you are thinking it, it is answered here.

No. The PAS stays. We work on the data layer around it — the warehouse, actuarial marts, regulatory reporting and analytics. Modernising the data estate does not require replacing the system of record.
Automated validation on every claim and policy object, with source and target running in parallel until reserves reconcile exactly. A discrepancy blocks the wave rather than being noted and passed on.
Yes, and this is the acceptance test we hold ourselves to. Model inputs are versioned with the data, so any prior valuation re-runs and produces the same number. If it does not, the migration is not finished.
Yes. 837 and X12 EDI parsing is already in production, and layout-aware parsing handles PDFs and scans through OCR. Most carrier data worth migrating is not in a tidy relational table.
The stack deploys in your own cloud account or on-premises. Where jurisdiction requires it, policyholder data never leaves your perimeter at any point in the engagement.
A two-to-four week assessment covering source inventory, data quality and a costed roadmap against your fixed regulatory dates. The output is yours whether or not you continue with us.

Set the objective. We will scope it honestly.

Bring us the estate as it actually is, deadlines included. Two to four weeks later you have a costed roadmap you own — whether or not you continue with us.