Observability 2.0: Shifting Left From Reactive Mon...
summary
Panel describes moving from reactive monitoring to a unified OTel-first stack that standardizes telemetry across 1000+ microservices using CNCF pieces like OpenTelemetry and Argo, then layering GenAI for intelligent sampling, anomaly pipelines, and AI-augmented root-cause analysis. Claims include reducing false alerts via pattern learning, shift-left instrumentation in the SDLC, building a company knowledge graph to make LLM reasoning reliable, and using bounded automation plus quotas and guardrails to cut MTTR and control token and ingest costs.
AI-written summary · about
chapters (10)
- 0:00Observability 2.0 introduction
- 1:06Traditional monitoring challenges
- 3:31Signal versus noise in observability
- 6:19Developers avoid early telemetry
- 9:29GenAI in observability engineering
- 13:00Making GenAI deterministic
- 17:53Reducing MTTR with AI
- 21:04AI for incident prevention
- 26:00LLM cost optimization in observability
- 30:33Legacy monitoring habits
related talks
29mObservability; a gentle introduction
11mObservability 2.0 - Much more than just logs,metrics and traces
39mShifting Left on Production Excellence with Observability
33mBeyond Monitoring: The Rise of Observability Platform
01mObservability
48mAPI Observability by Design: Building Systems That Tell Their Own Stories
13mObservability Where are We Headed?
50mPanel: Observability