My research experience is threefold. I currently focus on developing data driven model reduction algorithms and deep learning to capture the essential dynamics of atmospheric turbulence. In this regard I develop deep convolutional neural nets to predict extreme weather events/blocking events via image recognition of spatio-temporal continuous global circulation patterns. Along with this we are interested in lead time prediction via Conv-LSTMs. Previously, over the last couple of years I have worked extensively on High Performance Computing architectures , the performance portability of large scale linear algebra applications on multi-core and hybrid architectures aimed at multi-physics modeling. Previous to that I was involved in algorithmic development of computational geometry to model dynamic fluid interfaces under varying boundary conditions. Check out my research blog for short descriptions of a few things I had worked on. Most of my repositories are on my Github while one is private. Drop me an email if you want to use my libraries and need some help.