º«¹úÂãÎè

The complexity of modern cloud platforms and the highly distributed nature of today’s software systems present significant challenges for debugging, anomaly detection, and root cause analysis. Traditional methods for software debugging, monitoring, and performance analysis are often limited, as they focus primarily on known issues. The rise of large language models (LLMs) has further transformed software development by automating tasks, often reducing developers’ direct engagement with code and decreasing system awareness. This lack of awareness introduces new challenges, particularly in maintaining security, performance, and reliability. This talk will explore the concept of software observability and discuss mechanisms that support debugging, fault diagnosis, anomaly detection, and AIOps. It will highlight examples from ongoing research in the telecom domain and examine how observability can address the gaps in system understanding created by LLM-assisted coding. The discussion will conclude with an analysis of broader challenges and research opportunities in advancing observability for the modern era.

Speaker

Naser Ezzati-Jivan, PhD, is an Associate Professor in the Department of Computer Science at Brock University, Ontario, Canada. His research focuses on Software Performance Engineering and Software Analysis, supported by over 16 years of experience as a professor, software engineer, and team leader. He has collaborated on several research projects with major industrial partners, including Google Montreal, Ericsson, and Ciena. An active member of the academic community, Nazer has served as a reviewer and program committee member for major journals and conferences. He completed his PhD at Polytechnique Montréal, where his thesis was recognized with a special mention award.