AIOps
summary
AIOps frames observability as a data problem and proposes a five-step path: combine IT with data science, ingest logs/metrics and even social media, store everything in data lakes and machine learning pipelines, apply ML to surface actionable incidents from thousands or millions of events per second, and automate responses up to self-healing systems. Vendors (Nagios, Splunk, Datadog, etc.) already add AIOps features, and the claim is that more data improves ML accuracy and enables automated remediation.
AI-written summary · about
chapters (12)
- 0:00Start
- 0:43AIOPS & DEVOPS
- 1:08DEVOPS PRINCIPLE: SEE PROBLEMS AS THEY OCCUR
- 3:26OBSERVABILITY IS... Data
- 4:26DEVOPS PRINCIPLE: AUTOMATE EVERYTHING YOU CAN
- 5:41WHAT DO WE MEAN BY "AI"?
- 8:12GARTNER'S AIOPS INFORMATION FLOW
- 10:54AIOPS ADOPTION - DATA SCIENCE CAPABILITIES
- 12:29AIOPS ADOPTION-EXISTING & NEW FEEDBACK
- 13:50AIOPS ADOPTION - AIOPS TOOLS
- 16:01AIOPS ADOPTION - INTELLIGENT NOTIFICATIONS
- 18:01AIOPS ADOPTION - SELF-HEALING SYSTEMS
related talks
33mAge of AI Ops
29mAIOps Is How We Will Survive DevOps
02mAIOPS – How can machine learning help in IT operations
20mAutomated Incident Management
30mIntelligent Analysis on Cloud Infrastructure With AIOps
05mLightning Talk: AIOps vs. MLOps: What's the Difference?
27mAlert Fatigue to Action: Observability as the Backbone of Practical AIOps with OpenSearch
42mAIOps for Distributed Environments - Deep Dive