Dr Alim Ul Gias

Profile

I am Alim Ul Gias, a Lecturer in Programming at Queen Mary University of London. I hold a PhD in Computing from Imperial College London and have nearly a decade of academic experience across institutions in both the UK and Bangladesh. Before joining Queen Mary, I was a Lecturer in the Department of Computer Science at City, University of London, and a Research Associate at the Centre for Parallel Computing, University of Westminster.

Earlier in my career, I served as a Lecturer at the Institute of Information Technology, University of Dhaka, where I also completed my M.Sc. in Software Engineering and B.Sc. in Information Technology. As part of my undergraduate studies, I interned at Grameenphone. My research broadly focuses on software engineering and distributed systems, with a growing interest in computing education. More details are available on my Research page.

Teaching

IOT452u Software Engineering Tools, Techniques and Practices
IOT529u Algorithms and Data Structures
IOT554u Software Development Methods and Quality Assurance in Software Industry

Research

Research Interests:

My research lies at the intersection of software engineering, performance optimisation, and distributed systems. A major focus is on intelligent performance and reliability management in cloud and edge computing—for example, developing controllers that integrate predictive workload forecasting with reinforcement learning to make proactive scaling decisions based on real-time telemetry. These systems aim to optimise resource usage while meeting performance and availability targets, even under partial observability. They can be adopted as modular open-source tools across industry and the public sector, contributing to goals such as net-zero computing and smart manufacturing. A central theme of this work is addressing the unique challenges of cloud-native applications, particularly microservices, where distributed architectures and DevOps practices create complex trade-offs among performance, reliability, and cost.

I also explore how performance engineering can be more effectively integrated into software engineering education. Drawing on my experience with research tools and teaching distributed systems, I investigate AI-assisted learning resources that help students reason about system performance and reliability. This includes modular content, outcome-driven assessment frameworks, and adaptive lab exercises that simulate performance pitfalls and provide personalised feedback. The broader aim is to bridge the gap between theoretical understanding and practical expertise in modern computing environments.

Publications

A. U. Gias, Y. Gao, M. Sheldon, J. A. Perusquía, O. O’Brien, and G. Casale, “SampleHST-X: A Point and Collective Anomaly-Aware Trace Sampling Pipeline with Approximate Half Space Trees,” Journal of Network and Systems Management, vol. 32, no. 3, pp. 1–38, 2024.

A. U. Gias, Y. Gao, M. Sheldon, J. A. Perusquía, O. O’Brien, and G. Casale, “SampleHST: Efficient On-the-Fly Selection of Distributed Traces,” in Proc. of the 36th Network Operations and Management Symposium (NOMS), Miami, USA, pp. 1–9, IEEE/IFIP, May 2023.

A. Alnafessah, A. U. Gias, R. Wang, L. Zhu, G. Casale, and A. Filieri, “Quality-Aware DevOps Research: Where Do We Stand?,” IEEE Access, vol. 9, pp. 44476–44489, 2021.

A. U. Gias and G. Casale, “COCOA: Cold Start Aware Capacity Planning for Function-as-a-Service Platforms,” in Proc. of the 28th Int’l Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS), Nice, France, pp. 1–8, IEEE, Nov. 2020.

A. U. Gias, G. Casale, and M. Woodside, “ATOM: Model-Driven Autoscaling for Microservices,” in Proc. of the 39th Int’l Conf. on Distributed Computing Systems (ICDCS), Dallas, USA, pp. 1994–2004, IEEE, Jul. 2019.