Santosh Pande’s area of research is  compiler analysis and optimizations  which provide deep insights into the intricate behaviors of software and such insights are invaluable to improve multitudes of software systems properties in a highly automated manner.  He has done work both on static  as well as runtime analysis that tracks emergent program properties, the end result being optimizations as well as tracking the security properties during execution. Some of the highlights of his past research involve: (a) solving the challenging problem of operating system-less scheduling on very high performance throughput oriented (network) processor (Embedded Systems), (b) exposing a key vulnerability in XOM secure processor through a side channel attack and proposing a practical fix for it (Hardware Security) and (c) developing a systematic framework to understand the soft real time properties of modern (interactive) software converting an art into science and (4) utilizing algorithmic properties of software to produce highly scalable performance for computationally intense SAT solvers which are extremely demanding in terms of performance but can not be parallelized. He is applying this work currently to many ML algorithms. His work is disseminated through 100+ papers, open source software and has been consistently funded by NSF and is supported by ONR, DARPA and many companies.