Auto Insurance Warranty Claims

Overview

Insurance claims processing for automotive warranties with fraud detection indicators. This dataset contains detailed claim lifecycle events across multiple dealers and regions.

Download: auto-insurance-warranty-claims.csv (214 MB)


Dataset Summary

Metric Value
Events 685,702
File Size 214 MB
Process Type Insurance Claims Processing

Columns

Column Description
CaseId Unique claim identifier
ActivityName Claim processing step
ActivityTime Timestamp of the activity
Resource Person or system performing the activity
ClaimNumber Insurance claim reference
VIN Vehicle Identification Number
ProductType Warranty product type
VehicleMake Vehicle manufacturer
VehicleModel Vehicle model
VehicleYear Vehicle year
Mileage Vehicle mileage at time of claim
ContractType Warranty contract type
DealerName Servicing dealer
DealerRegion Geographic region of the dealer
CustomerState Customer state
ClaimValue Dollar value of the claim
CustomerType Customer classification (e.g., Fleet)
PriorClaims Number of previous claims
PredictedFraudScore AI-predicted fraud probability
PaymentAmount Amount paid on the claim
Priority Claim priority level
ApprovalLevel Required approval level
DiagnosisCode Repair diagnosis code
FraudIndicators Flagged fraud indicators
LaborHoursClaimed Labor hours claimed by dealer
LaborHoursApproved Labor hours approved
PartsValue Value of parts used

Suggested Analyses

  1. Process Map: Visualize the claim processing workflow from first notice to payment
  2. Fraud Detection: Filter by high PredictedFraudScore to identify suspicious patterns
  3. Dealer Performance: Compare processing times and approval rates across dealers and regions
  4. Labor Hour Disputes: Analyze cases where LaborHoursClaimed differs from LaborHoursApproved
  5. Cycle Time: Measure time from First Notice of Loss to final payment

This documentation is part of the mindzie Studio process mining platform.