Source & de-identify
Claims are sourced under license and stripped of PII through a deterministic de-identification pipeline with k-anonymity checks.
SEC-03Data Quality & Verification
Verified like the claims themselves
Every record passes a four-stage verification pipeline run by licensed adjusters, forensic specialists, and automated leakage controls.
Expert annotation
Licensed adjusters and forensic specialists label damage, causation, coverage, and fraud indicators under calibrated rubrics.
Adjudication & consensus
Every gold-standard record is reviewed by two independent experts; disagreements are resolved by a senior adjudication panel.
Benchmark validation
Records are stress-tested against ClaimBench task suites, checked for leakage, and versioned with full provenance metadata.
κ = 0.94
Inter-annotator agreement on gold-standard sets
100%
PII de-identification coverage, independently audited
0
Train/test contamination events across all releases
Contact Sales
Talk to our data team
Tell us about your models and the claims data you need. We respond to every serious inquiry within one business day.
- Dataset samples shared under NDA within 48 hours
- Enterprise licensing with on-prem or VPC delivery
- Custom generation scoped by our solutions engineers