Using AI to generate YAML is one thing. Letting it touch production operations is another.

YongKang He says the safest starting point is low-risk, high-volume work like anomaly detection, correlation, and remediation suggestions. He is not ready to hand over costly scaling decisions or sensitive policy changes without stronger guardrails.

The practical lesson is that AI should act like a co-pilot for SRE, not a fully autonomous operator.



Watch the full interview: https://ku.bz/8Q7Vy60P7