U.S. power consumption is surging after decades of relative stability driven by AI, data centers, and manufacturing. Furthermore, extreme weather events are increasing in frequency and severity, stressing aging power infrastructure and threatening grid resiliency. Thus, optimal infrastructure planning is critical for power systems to meet future electricity demands. Stochastic programming is a well-suited optimization paradigm for modeling uncertainties in energy loads and renewables, but it can be computationally intensive. This talk presents SPAROW: an open-source Python library for stochastic programming, distinguished by its support for scenario bundling, multifidelity modeling, and a rich solution pool manager. SPAROW leverages decomposition techniques, including Progressive Hedging, Benders decomposition, and spatial branch-and-bound, to accelerate computation by solving subproblems in parallel. We provide a power systems planning exemplar to demonstrate the efficacy of our approach.