Patagonia is one of the key places to study the interaction of plate tectonics and mantle flow patterns with geological processes. This part of the continent is shaped by the northward migration of the Chile Triple Junction, currently marked by subduction of the Chilean spreading ridge at latitude 46oS, opening a slab window beneath Southern Patagonia. The idea of slab window was hypothesized to explain the volcanic gap between north Patagonia and the southern part of the peninsula. The analysis of volcanic rock composition shows the transition between a domain with the signature of slab melt (metasomatized MORB) and a domain with no slab signature (OIB source mantle). Along the Pacific coast, other slab windows were suggested in Central America, California and North Cordillera. The analysis of uplifted terranes and seismic imaging tried to constrain the geometry of these slab windows and map the mantle flow pattern that controls the present-day surface expression (topography, volcanism distribution). From a limited seismic coverage, early studies mapped the Patagonian slab window from body wave tomography and shear wave splitting. The recent deployment of a temporary seismic array from 2018 to 2021 and the Chilean seismic networks fills the data gap between the seismically active northern part of Patagonia and the more poorly studied southern part. This presentation will show the results of our recent seismic studies in Patagonia and help constrain the geodynamical processes associated with the slab window. From the analysis of SKS and similar core phases, we determine the pattern of azimuthal seismic anisotropy resulting from the mantle flow pattern beneath South America. Fast splitting directions are generally NE-SW throughout most of Southern Patagonia, similar to the pattern of large-scale azimuthal seismic anisotropy from global and regional surface wave models. However, between 46oS - 48oS, we observe large splitting values and an E-W direction showing the effect of the slab edge. This is consistent with models of rapid upper mantle flow from the Pacific around the southern edge of the Nazca slab. Seismic imaging using receiver functions and Rayleigh waves from earthquakes and ambient noise show very low upper mantle velocities and an absence of mantle lithosphere in this region, suggesting the lithosphere has been thermally eroded by the dynamics of the slab window. We will also show and discuss preliminary results of a body wave tomographic analysis of the same seismic station dataset.