Passed the 100+ document release gate.
Loss Run
Claims
loss_run
Claim ledger extraction for policy periods, claim numbers, paid and reserved amounts, statuses, and loss-history evidence.
Available through API, processor, playground, CLI, and MCP.
Schema family used for routing, docs, corpus, and benchmark grouping.
Part of the full 56-schema 99%+ accuracy release.
API Surface
Every schema has the same inspection rails.
Developers and agents can inspect the schema contract, corpus state, benchmark state, and schema memory before choosing whether to call extraction.
GET /v1/schemas/loss_run GET /v1/corpus/loss_run GET /v1/benchmarks/loss_run GET /v1/benchmarks/loss_run/latest GET /v1/schema-memory/loss_run
Values need a reason.
Returned fields must tie back to source text, page or section references, evidence quality, and extraction provenance.
No proof, no truth.
Weak, missing, or unsupported evidence stays unresolved instead of becoming a plausible JSON value.
Promotion passed.
Every public schema became release-ready only after its 100+ document corpus passed required-field, evidence, exact-label, and document-pass scoring.
Related
More in Claims.
hardened
First Notice of Loss
First Notice of Loss extraction for loss dates, claimant names, loss locations, and claim descriptions from real public claim notice records.
fnolhardenedClaim Packet
Claim packet extraction for claim numbers, loss facts, coverage references, open items, correspondence, and supporting exhibits.
claim_packethardenedAdjuster Report
Adjuster report extraction for damage summaries, coverage positions, claim references, inspection notes, and recommendation evidence.
adjuster_reporthardenedRepair Estimate
Repair estimate extraction for vendors, labor, materials, line items, totals, and estimate-supporting damage descriptions.
repair_estimatehardenedMedical Bill
Medical bill extraction for providers, service dates, charges, diagnosis references, and claim-related billing evidence.
medical_bill