This dataset provides structure-level assessments for buildings across the United States, using recent aerial imagery and historical loss experience to model risk for underwriting, pricing, and reinsurance. Each property record includes dozens of structural characteristics such as roof condition, geometry, age, and nearby vegetation, combined with peril models for severe convective storms, wildfire, water-related loss, and attritional fire. Outputs express the likelihood and severity of loss at the individual structure rather than aggregated territory grades. Spatial identifiers enable mapping, proximity analysis, site selection, portfolio aggregation, and regional market evaluation. The data supports deep exposure analysis, automated underwriting workflows, and reinsurance modeling.

Key attributes include:

• Structure-level physical features including roof and building condition and geometry

• Peril-specific risk measures expressing probability and expected severity of loss

• Recent high-resolution aerial imagery as the primary observational source

• Historical loss and claims data integrated for model calibration and validation

• Spatial identifiers supporting mapping, proximity and portfolio aggregation analysis

Value Summary:
Insurers, reinsurers, MGAs, lenders, and risk managers benefit from granular, evidence-driven assessments to improve underwriting, pricing, portfolio selection, and exposure management. The dataset reduces reliance on coarse territory averages and enables property-specific decisioning and loss forecasting.