Large-scale disasters such as earthquakes, tsunamis, hurricanes, and other windstorms reveal how vulnerabilities in the built environment manifest as severe economic, environmental, and societal impacts, which threaten the lives and livelihoods of communities worldwide. The effective reduction of these disaster-related losses requires the development of targeted mitigation strategies and policies informed by faithful regional loss assessments under realistic hazard scenarios. Such capabilities are now being advanced by investments in a new generation of open-source loss modeling workflows that evaluate hazard impacts on actual communities. While these open-source workflows do provide the parcel-level granularity stakeholders now require, hurricane-related losses in these workflows are still projected using non-specific characterizations of each building in the inventory. However, the realization of site-specific, building-specific, and component-level hurricane regional loss assessments requires the ability to support evaluations for the diversity of geometries and material assemblies found in today's building stock; this presents a significant data challenge. Fortunately, today's open data landscape creates new opportunities to gather the data necessary to enable more granular, high-fidelity descriptions of hurricane impacts on actual constructed buildings.In response, this dissertation capitalizes on this newfound, open-data landscape to formalize methodologies that (1) deepen the capacity for high fidelity loss assessments of actual constructed buildings under multiple hazards through an integrated life-cycle assessment (iLCA) framework, (2) establish computationally-efficient, data-driven frameworks for hurricane regional loss assessments that (i) harness the modern open building data landscape, (ii) support the myriad of geometries and material assemblies found in today's building stock, and (iii) facilitate seamless scale up from parcel to regional scale, and then (3) implement these frameworks using actual constructed buildings/building inventories to demonstrate how open data can be leveraged to advance various modules of existing end-to-end open-source hurricane regional loss assessment workflows. The dissertation's integrated life-cycle assessment framework deepens the capacity for high-fidelity loss assessment of actual constructed buildings under multiple hazards. The resulting end-to-end workflow's ability to realize site-specific, building-specific, and component-level hurricane regional loss assessments is then facilitated by the dissertation's development of a series of computationally-efficient, data-driven frameworks bolstering specific modules in the workflow. A new building data model supports automated regional loss assessments and capitalizes on publicly available data. This model serves as the scaffold for a data-driven framework enabling the automated simulation of the site-specific effects of wind and WBD hazards on actual constructed buildings and a Bayesian data integration framework to develop component-level fragilities using damage observations from post-disaster datasets. These capabilities are exercised on individual buildings and then scaled up through the replication and extension of existing open-source workflows for hurricane regional loss assessment. This is made possible by using the dissertation's methodology for generating building inventory information using open data and heuristic rulesets to capture parcel-specific, time-varying attributes and enable an immediate expansion of the capabilities available in today's open-source loss modeling workflows. All of these data-driven advancements are implemented using real-world building inventories, prioritizing considerations of commercial building typologies located in Florida's Bay County, the landfall site of Hurricane Michael. The selection of this region enables validation of the dissertation's findings against observations in recent hurricanes, while also grounding data-driven advancements within the current availability of tax assessor data and other publicly available datasets.