Cloud Convergence Summit

Multi-cloud Kubernetes environments increase operational entropy with more network paths, greater latency variance, and a larger failure surface. Traditional monitoring surfaces symptoms but does not govern system behavior. When automation or AI-assisted optimization is introduced without safety boundaries, it can amplify instability rather than reduce it.

Tune into this session from Walmart Global Tech’s Sibasis Padhi as he presents a practical approach to closed-loop reliability governance built around percentile-based SLO enforcement (p95/p99), breach-triggered optimization, corrective actions, and measurable before/after validation. The focus is vendor-neutral and implementation-oriented, showing how cloud and platform teams can move from reactive monitoring to controlled, auditable optimization across distributed environments.

Key Takeaways:
- Why p95/p99 latency; not median - should drive operational decisions.
- How SLO breach detection becomes the trigger for safe automation.
- Guardrails that constrain optimization actions (bounded change, rollback readiness).
- A practical checklist for implementing reliability governance on Kubernetes.