I am currently a research Associate at Centre for Parallel Computing, University of Westminster. I completed my PhD. at Imperial College London, under the supervision of Dr. Giuliano Casale. Before that, I was a lecturer at Institute of Information Technology (IIT), University of Dhaka (DU). I completed M.Sc. in Software Engineering from IIT, DU, with thesis on Adaptive Software Performance Testing. I received my bachelor’s degree in Information Technology (major in software engineering) from the same institute. As a part of my bachelor's program, I also worked at Grameenphone as an intern. My primary research interests are QoS and resource management of cloud-native applications.
My current research focuses on different Quality of Service (QoS) aspects of cloud-native applications e.g., microservices. Due to its synergy with DevOps, cloud-native applications are quickly getting adopted in different software development companies. However, they introduce a lot of performance and resource management challenges due to their unique architectural pattern. Besides, there are challenges involving reliability due to their inherent highly distributed nature. Therefore, rather than only focusing on performance, we might need to consider both performance and reliability. My primary focus is to design runtime capacity management processes that can ensure both availibity and performance constraints.
I am also interested in Automated Software Engineering. To be more specific, I like to address the challenges which are involved in automating any steps of the Software Development Life Cycle (SDLC). During my master's thesis, I worked on adaptive software performance testing. The main purpose of this research was to automatically suggest performance hotspots within a large scale web application. To do so, we frequently needed to use the concepts of metaheuristics . Such metaheuristic approach can also be useful in different software design problems involving optimization. I find such problems interesting and worked on one similar problem where the goal was to automatically suggest a component based design (having high cohesion and low coupling) from legacy code written in procedural language.