Live Q&A Session #1
summary
A live panel walks through concrete practices: SLO tooling gets tailored to decision personas with dashboards that show uptime versus error windows and concrete rollback or vendor-change actions; a production-trace‑derived service dependency graph produces filterable call graphs so engineers find failing hops quickly; teams build agentic frameworks that combine RAG with tool invocations and enforce a feedback loop to fix weak troubleshooting guides; recoverability follows a ring model and an ROI‑first path, and AI outputs are labeled recommendations used only for diagnostics not automated destructive fixes.
AI-written summary · about
chapters (12)
- 0:00Reliability at scale Q&A
- 0:25Service level objectives
- 2:29Service dependency graphs
- 4:17Standardizing data for AI
- 6:22Data gathering for AI
- 7:45Reliability spectrum
- 11:55Adding SLOs after launch
- 13:21Recoverability in dependencies
- 15:08Engineering reliability efforts
- 17:35Reliability program maturity
- 19:19AI agent guardrails
- 20:45Closing remarks
