Every claim decided on the page behind it.
Seven checks from first notice to recovery. Each cites its page and seals a signed receipt.
Seven checks. Each knows when to stop.
Gothink AIR reads the claim file. Conflicts between documents are kept and cited.
No step can deny a claim — a person closes it. Every threshold is a starting point for your claims, SIU and actuarial teams.
Paid without a person only once earned.
Granted per claim type, from reviewed outcomes — the same autonomy model as the rest of the platform.
The number of reviews is set by your error budget; a stricter budget needs more.
You choose the claim types
With no configuration, nothing is paid unattended.
Reviews come first
The console shows how many clean reviews are still needed.
One override revokes it
Unattended payments are sampled for QA.
The claim file, as it arrives.
The claims pack covers:
Reads the clauses
An exclusion becomes a question citing clause, page and sentence.
A cited chronology
Encounters, codes, gaps. Treated as PHI, shown only to authorised users.
Documents find their claim
Late documents are attached and flagged, not re-decided.
One console for the desk. One trail for the examiner.
Claims queue
Accept, modify, override or escalate — by a named reviewer.
Autonomy view
Reviews, errors and clean reviews still needed, per claim type.
Actions outbox
Every payment recorded before it is sent, marked sent or failed after.
Measured before it is trusted.
| Piece | What it does | Status |
|---|---|---|
| Guidewire ClaimCenter | Payment, referral and note — with a dry-run check that sends nothing | Template |
| Duck Creek Claims | Payment, task and note, checked the same way | Template |
| Subrogation pilot kit | Recovery screen on your closed claims. Nothing leaves the machine. | Built |
| Claims console | Queue, autonomy, outbox, subrogation backtest | Cloud · demo data |
Templates are refused until verified against a carrier's own tenant.
What is not built yet.
A verified core-system tenant
Verified against a carrier's tenant, not ours.
Litigation propensity
Only trained on a carrier's closed claims with outcomes.
An accuracy figure
The demo uses synthetic claims. It shows behaviour, not accuracy.
Start with your closed claims.
Run the subrogation screen inside your environment, then decide whether to go further.