Arnak Poghosyan

Distributed Tracing for the Troubleshooting of Native Cloud Applications via Rule-Induction Systems


Arnak Poghosyan has Ph.D. in Mathematical Modeling, Numerical Methods, and Software Complexes with extensive experience in academic and industrial research domains. His research is related to approximations and interpolations, numerical methods, and ML applications. Inventor of 70+ US patents in AI operations for automated cloud management.

Description of the Talk:

Diagnosing IT issues is a hard problem in a large-scale distributed cloud environment. Modern monitoring systems rely on ML/AI-empowered data analytics for detection, root cause analysis, and accelerated remediation of performance degradation. However, the successful adoption of AI solutions is anchored on trust. System administrators will not blindly follow the recommendations of AI analytics without sufficient interpretability of solutions. Explainable AI is gaining more and more popularity by enabling improved reliability, confidence, and trust in smart solutions. In this paper, we show the benefits of rule-induction classification methods, particularly RIPPER, for the root cause analysis of performance degradations of native cloud applications. It reveals the causes of performance degradation in a set of rules which administrators can use in remediation processes.