With drastic improvements in machine learning capabilities over the recent years, practical AI solutions have been deployed at scale in production-ready scenarios. In the landscape of observability, major commercial platforms have provided their users with various AI-enabled functionalities, mostly anomaly detection, since 2015. At the Apache SkyWalking ecosystem, open-source developers and young researchers are working on an community-driven AIOps solution to lower the bar for the curious practitioners.

As a young open-source initiative, focusing on reliable reactive anomaly detection is the key to building a concrete basis for later phases. Therefore, the project scope in phase one is defined as “To provide a pluggable AIOps anomaly alert engine with out-of-box integrations to popular observability platforms like Apache SkyWalking (Full-Stack APM) and Prometheus (Metrics Systems).” This presentation will introduce the basic concepts of AIOps for observability, the definition, justification and the design of the project.

Speakers:

Chen Yihao: Queen’s University, Master’s Student, Apache SkyWalking Committer, Yihao Chen is a current master’s student at the Software Analysis and Intelligence Lab (SAIL) at Queen’s University and a Committer at Apache SkyWalking, he focuses on community building and emerging technology adoption.