Both coastal and inland flooding affect U.S. communities each year, causing substantial damage to the built environment and long-term socio-economic disruption. To support proactive risk mitigation, this study develops a generalized and scalable framework for building-level, resilience-informed flood risk assessment. The framework integrates FEMA's National Flood Hazard Layer, fragility-based vulnerability functions, social vulnerability indices, and community resilience factors. A geospatial algorithm is developed to generate coastal and inland flood hazard maps for any U.S. community using FEMA flood data. Building-specific attributes from the National Structure Inventory are used to assign minimal flood archetypes and corresponding depth-damage fragilities to diverse building typologies, enabling high-resolution damage assessment in terms of exceedance probabilities of prescribed damage states. Social vulnerability and community resilience indicators from FEMA's National Risk Index are incorporated to account for unequal social burdens and adaptive capacity. Machine-learning classification techniques are then applied to categorize buildings into risk-based groups for mitigation prioritization, planning, and investment. Two complementary metrics are produced: a Physical Flood Risk Score reflecting risk of structural damage only, and a Resilience-Informed Risk Score that integrates physical, social, and resilience dimensions into a unified index of community-contextualized flood risk. The framework is demonstrated through a case study of North Carolina using 100-year flood hazard maps at building, census-tract, and county scales.
