Alert Fatigue to Action: Observability as the Backbone of Practical AIOps with OpenSearch
summary
Argues that unifying logs, metrics and traces into OpenSearch plus OpenTelemetry and Prometheus turns observability into an AIOps substrate that actually closes the loop from detection to fix. Describes indexing deployment metadata and runbooks, a workflow that dedupes and triages alerts, then invokes HolmesGPT for root cause, and finally opens GitLab merge requests to remediate. Shows integration via MCP and Cloud Code skills, mentions a live setup with KeepHQ, ~120 connectors, and a quick agent prototype that performs queries and actions.
AI-written summary · about
chapters (12)
- 0:00Alert fatigue and AIOps
- 2:05Managing production alerts
- 3:50CI/CD with OpenSearch
- 7:00OpenSearch observability stack
- 10:16OpenSearch and AI tools
- 13:51Root cause with Holmes GPT
- 16:18Deployment context for incidents
- 18:27Reducing alert noise
- 20:14Cloud Code for AIOps
- 23:25Future of incident response
- 25:16Live demo: Keep HQ workflows
- 26:48Q&A
related talks
42mObserve Smarter, Not Harder: Scaling AI-Powered Observability With O...
06mSponsored Keynote: OpenSearch - See Everything: Open Observability for Agentic AI
37mBuilding Modern Observability Architecture With OpenSearch, OS Ingest...
22mFrom Telemetry to Insight: Building AI-Powered Observability Pipelines with OpenSearch
29mObservability; a gentle introduction
20mGenAIOps - OpenSearch for AI & GenAI Platform Observability
41mCustomer Centric Observability at Scale Leveraging AIOps & OpenTelemetry
31mCharting New Frontiers: The Evolution of Observability in Cloud Environments