InsuranceStructured + TextAvailable
Insurance Claims Corpus: Property and Auto
Annotated property and auto insurance claims from an InsurTech platform. Covers claim narratives, adjuster notes, settlement outcomes, and fraud flags. Structured fields paired with free-text narratives, ideal for claims AI and fraud detection models.
Dataset Highlights
1.8M claims across property, auto, and liability lines
Fraud flag labels on 420K claims
Settlement amount and disposition metadata
Adjuster note text with loss codes
Dataset Specifications
Volume
1.8M claims
Format
Parquet + JSONL
Date Range
2015–2023
Token Count
~2.4B tokens
Collection Method
InsurTech platform export with policyholder consent
Tags
InsuranceClaims AIFraud DetectionStructured
Dayda Verified Listing
This dataset has passed Dayda's vetting process, with legal provenance reviewed, data ownership confirmed, PII screened, and quality spot-checked. The seller's identity is only revealed after an NDA is signed.
Price Range
$90,000–$180,000
Final price subject to licensing terms and negotiation.
Licensing Options
non-exclusiveperpetual
Dayda Quality Score
90/100
A standard NDA is required before samples are shared.
What Happens Next
1You submit your interest and use case
2Dayda sends a standard NDA
3You receive a 1–5% data sample
4Submit an offer. Dayda handles negotiation
5Data Purchase Agreement executed
6Secure data transfer + payment