DevOps at LLM Speed: Using an AI Copilot for Kubernetes and Jenkins
summary
Claims a 50% cut in incident response time at Storylane by using Skyflo.ai, an open-source, self-hosted AI copilot that links Kubernetes, Jenkins, Argo, Helm and CI/CD. Demo triggers a Jenkins build, watches the pipeline succeed while Kubernetes pods fail, and shows the agent running discovery, creating a plan, asking for human approval before mutating, then verifying outcomes. Under the hood: a custom MCP server with 50+ typed tool calls, LangGraph workflow, and streaming via Redis/SSE; audit logs and Apache‑2 license.
AI-written summary · about
chapters (10)
- 0:00DevOps AI automation
- 1:59Toolchain challenges
- 3:24Skyflow and postcode bottleneck
- 5:34Demo: Skyflow AI copilot
- 16:02Production safety design
- 22:45Skyflow architecture
- 23:55Self-hosting and data ownership
- 25:16CI/CD and human-in-loop
- 28:15Authentication and audit logs
- 30:14Multi-agent vs unified agent
related talks
44mAgentic Delivery, Guardrailed: AI in CI/CD & Platform Engineering (Open Source Patterns)
41mEfficient DevSecOps workflows with a little help from AI
42mLLMs on Autopilot: Running AI Agents on Kubernetes With Open Source Tools
29mThe Future of DevOps: AI, Security, and Next-Gen Collaboration
26mAgentic DevOps in the Autonomous Cloud
12mThe Synergistic Impact of Artificial Intelligence on DevOps
22mUnlocking AI Powered DevOps Within Your Organization
39mThe evolution of Devops in the age of AI