An incoming demand, fact-checked against the actual record.
Demand package audit software checks every diagnosis, treatment claim, and billed figure in an incoming demand against the medical records behind it — discrepancies flagged before a counter gets calculated. Your team still decides the counter; the audit supplies the cited facts, each one traced to its source page.
Every assertion, checked against its own exhibits.
The demand letter tells a story; the enclosed records are the evidence. The AI reads both — every diagnosis claimed, every treatment course described, every causation statement — and matches each assertion to the pages that should support it. What the exhibits actually show is the finding.
Billed figures, reconciled to the ledger.
Claimed specials are checked against the billing documentation enclosed with the demand: session counts against the visit notes, totals against the itemized bills, and duplicate charges surfaced along the way. Where the demand's number and the record's number differ, you see both, with a page reference for each.
Discrepancies flagged, evidence attached, humans decide.
The audit's output is a flag list: what the demand asserts, what the record shows, and where both live. Your adjuster or counsel walks into the negotiation with the discrepancies documented — the counter itself, and every claims decision behind it, stays with your team. The software never prices a settlement.
Every flag cites both sides.
A discrepancy you can't source is an accusation; a discrepancy with two page citations is leverage. Every finding in the audit is source-linked and legally defensible — the demand's line and the record's page, side by side. Where the record is silent, the audit says "not documented" rather than guessing.
See Verifiable AI CitationsFrom demand package to flag list.
Three steps, before anyone drafts a number.
The letter plus every enclosed exhibit — records, bills, imaging reports — in one drop.
Diagnoses, treatment claims, and billed figures each return supported, unsupported, or contradicted — cited.
Your adjuster reviews the discrepancies, verifies any flag in one click, and builds the counter on documented facts.
Who audits demands with it.
Anyone on the receiving end of a demand letter with exhibits attached.
Bodily-injury adjusters evaluate the demand against its own exhibits before reserving or countering — every discrepancy documented.
For carriersA consistent, cited audit on every incoming demand across clients — the same rigor at desk-load volume.
For TPAsDemand package audit, answered.
Upload the demand package and its supporting records together. The AI reads the demand's assertions — diagnoses, treatment history, causation statements, billed amounts — and checks each one against the medical records behind it. Every assertion comes back supported, unsupported, or contradicted, with page citations into both the demand and the record.
Yes. Billed totals and treatment counts claimed in the demand are checked against the billing pages and visit documentation actually enclosed: a claimed number of PT sessions is counted against the therapy notes, and claimed charges are matched to the itemized bills. Where the demand's figure and the record disagree, both numbers are shown with their sources.
Explicitly, item by item. Each discrepancy states what the demand asserts, what the record shows, and where both live — a diagnosis with no supporting note, a treatment period longer than the visits documented, a prior injury the demand omits. Nothing is buried in a summary paragraph; the flag list is the deliverable.
It feeds the human who calculates it. The audit hands your adjuster or counsel a verified-versus-flagged breakdown of the demand with citations attached, so the counter is negotiated from the documented facts. The software never computes a settlement value or makes the claims decision — evaluation stays with your team.
Yes. Settlement Demand Letter drafts an outgoing, citation-backed demand from the record; Demand Package Audit reads an incoming one and checks it against the record. Same engine, same citation discipline — pointed in opposite directions across the negotiating table.
Related capabilities.
The audit draws on the platform's billing and record intelligence.
The mirror image: an outgoing, citation-backed demand drafted from the record.
See demand draftingEvery charge in the file as one reconciled, cited ledger — the audit's billing backbone.
See billing ledgerHow the enclosed charges compare to benchmarks — context for the flagged figures.
See bill benchmarkingRoute the audited demand to the right desk — sorted by documented criteria, humans deciding.
See claims triageAudit a demand sitting on your desk right now.
Upload one demand package and get the supported-versus-flagged breakdown back, cited on both sides. Handled under our BAA; never used to train a model.