Claim intelligence
Every insurer in the world reads documents, reconciles evidence and decides claims. What differs - completely, and in ways that do not generalise - is what the regulator requires, what the customer is owed, and what a defensible decision has to look like. We build the engine once and the obligations per market.
One market shipped. Nine engines, 179 tests, zero dependencies. We would rather be genuinely deep in one jurisdiction than shallow in twelve.
The thesis
A system that reads a medical certificate works anywhere. A system that knows the certificate is a statutory instrument with a review cycle, that the clock it starts runs in business days under one state's calendar, and that stopping payments on it requires notice - that system only works where it was built.
Document understanding. Evidence anchoring. Cross-document reconciliation. Provenance and replay. Entity resolution. The architecture of an explainable decision.
Built once
Obligation timeframes and their units. Benefit formulas and indexed caps. What evidence a decision requires. Dispute pathways. Public holidays, of all things, which change the answer.
Built per market
Findings are never decisions. No assertion without two sources. Protected attributes firewalled from assessment. No opaque score anywhere in the product.
Everywhere, in code
The uncomfortable consequence is that entering a market is not a configuration exercise. It is reading the instruments, encoding them as data with citations, and having a person verify each one against the primary source. That is slow, and it is the moat.
Platform
Each answers one question well. The facts that decide whether a claim ends well sit between them, which is where a single-purpose tool cannot look.
An example of the last one: you are about to stop weekly payments on a claim that has been underpaid for ninety-seven weeks. The obligation engine does not know about the underpayment. The entitlement engine does not know about the cessation. Nobody in the process sees both.
Markets
A market is "live" here only when its obligations are encoded with citations and its benefit arithmetic is modelled and tested. Anything else would be a landing page pretending to be a product.
71 obligation rules across 8 jurisdictions. 4 benefit schemes. General Insurance Code, RG 271, APRA CPS 230 and 234, Privacy Act APP 1 and 11, state personal injury schemes.
See the Australian product →The closest adjacent market: Fair Insurance Code, FMA conduct licensing, and an accident scheme with no direct Australian analogue. Shares more architecture with Australia than any other market does.
FCA Consumer Duty puts outcomes and evidence at the centre, which suits this architecture unusually well. Different enough in claims practice that it is real work, not a translation.
Provincial regulation and provincial accident benefit schedules - structurally the same problem as Australian state schemes, which is why the engine already has the shape for it.
Consumer Protection Code and the EU AI Act's treatment of automated decisions affecting people. The transparency obligations here are the strictest anywhere, which is the case this product was designed for.
Fifty regulators, unfair claims practices acts per state, and a market where the incumbents already sell speed and fraud detection. We would need a reason to be there beyond size, and we do not have one yet.
If your market is not listed and you have the regulatory problem this solves, that is worth a conversation - the order we build in should be set by who actually needs it. [email protected]
What we will not do
No accuracy, detection-rate or time-saved figure appears anywhere on this site, because no pilot has produced one under a methodology we would publish beside it. When one has, the number and the method arrive together.
All 71 obligation rules and every indexed rate carry verified: false until a
named person has checked them against the primary instrument. The count appears in every
position and every report rather than in a footnote.
There is no risk score, no ranking of claimants, and no composite figure spanning compliance and evidence. The only number attached to a person is how many independent documents corroborate a fact about them.
The Australian product is live, documented and demonstrable on synthetic claim files. It is the fastest way to judge whether the architecture is right for your market.