Machine Learning is perceived as the way to sip through large amount of data, draw conclusion, capable of predict outages, pinpoint root cause, suggest remediation. After many years of trying, the result is still not ideal (“oh no, false positive!”). Does this mean we should give up? If not, how can we turn the “not so ideal” results into something useful? This session will clarify the ML myths and hypes. Share how machine learning can be used pragmatically; and how can clients help make this a reality. Collect feedback on a few approaches such as Event Grouping, Gen AI and LLM; brainstorm on other ways to make AI useful. Brainstorm on the balance between leveraging the power of AI and safeguarding wrong decision or actions in IT operation space.