Research

My research interest is in the general area of performance evaluation of software systems. Specifically, the research topics I am currently interested in are -

∙ Measuring, modelling and improving the performance of enterprise applications

∙ Performance characterization and modeling of Cloud and Internet of Things (IoT) systems

∙ Using Machine Learning algorithms to evaluate and improve software performance

Software performance is a crucial aspect of enterprise applications. In absence of proper software management techniques, applications will suffer from bad performance triggered by unidentified performance bottlenecks and will scale poorly due to the lack of proper capacity planning. A prime case in point is the colossal failure of the Obamacare website that prevented a vast majority of users to access the application when it was initially launched in 2013 caused due to the lack of proper performance management. In my research, I develop intelligent, dynamic, scalable, cost-optimized techniques for managing application performance. My research investigates automated runtime modelling techniques that can be used to manage the performance of applications such as cloud-based Web services, video streaming services and IoT enterprise applications.

I focus on research problems that are academically rigorous and has immediate practical relevance in the industry. To this end, I have a long history of academic collaboration with premiere industry research laboratories such as Hewlett-Packard (HP) Labs, Palo Alto, USA and IBM Center for Advanced Studies (CAS), Toronto, Canada. I have also collaborated with startup companies such as Predikat Inc., Ottawa, Canada.