While Navier-Stokes modeling (e.g., LES, RANS) is often needed for modeling the physics of turbulent flow close to a body, far-field wave propagation can often be considered inviscid, and Navier-Stokes solvers can be computationally expensive and too numerically dissipative to model waves over long distances. As a result, it is useful -- if not necessary -- to consider a hybrid approach ; in this talk, results will be presented for coupled viscous-inviscid computations, with applications to tsunami propagation and wave-impact to offshore structures (e.g., naval ships, offshore wind turbine foundations). Particular attention will be given to the solution of the inviscid flow with the boundary element method (BEM), based on fully nonlinear potential flow theory, and the developments necessary to improve the accuracy and speed required for large-scale problems. Examples showing the improved accuracy / speed will be shown, as well as the numerical and physical details to resolve for the hybrid approach.