DO WE STILL NEED TO "OBSERVE" THE FUTURE OF AI & OBSERVABILITY
summary
Claims that AI eliminates no PromQL by translating plain English into vendor queries, automates correlation and causal-finding, and runs SRE agents that draft rollback PRs and remediate automatically. Cites observability cost as 15–25% of infra spend and noisy alerts around 70%, and adoption of 78% using OpenTelemetry. Describes an AWS reference agent + GitHub sample, dynamic routing to hot/cold storage (Glacier), and auto-instrumentation; prototypes are runnable and aim to cut investigation time from tens of minutes to near-autopilot fixes.
AI-written summary · about
chapters (12)
- 0:00Introduction and talk overview
- 1:06Incident response pain points
- 3:40AI in the SDLC and OLC
- 6:39Observability stack breakdown
- 9:33Intelligent data collection
- 13:51OpenTelemetry and AI integration
- 15:34AI-powered query and insights
- 18:44Agentic monitoring setup
- 20:57SRE support agent design
- 23:07Portable observability stack
- 26:22Future of AI-driven observability
- 27:30Q&A and closing remarks
related talks
19mDo We Still Need to “Observe”? the Future of AI & Observability
28mWhere are we going with observability?
44mAgentic Observability — The Path to AI Co-SRE
08mThe Future of Observability: Trends, AI, and New Relic’s Vision f...
25mWhen Your "New Teammate" Hallucinates: Observability as the Safety Net for AI Code
11mObservability for the AI Cloud
28mImplementing AI-Driven Observability in Healthcare
33mWhere Are We Going With Observability?