Proposal: Challenges in Open-Source AIOps Platforms for Proactive Anomaly Detection with Minimal False Positives

[DanielBabenko]

Greetings, colleagues.

I am student from ITMO University and I am having the course “Cloud and Fog Computing” now. As a project for this subjects I have selected such topic / problem to solve:

“Challenges in Open-Source AIOps Platforms for Proactive Anomaly Detection with Minimal False Positives

Open-source AIOps tools exist, such as Keep and Seldon Core, but a significant deficiency remains in their ability to provide highly accurate, AI-powered anomaly detection that minimizes false positives, especially in dynamic cloud environments, to prevent alert fatigue."

My goal for the subject is to make a pull request to an open source project that closes a practical aspect of the topic. I am not quite experienced with anomaly detections, so I will appreciate if you can give me some advices on how I could improve this toolkit, so I could minimise False Positives while not breaking its logic.

Is this feature useful for Seldon Core?
If yes, would you mind to help me with its implementation by sharing dome piece of advice?

Thanks,
Daniel Babenko.