🚀 The End of YAML? Managing Infrastructure with NL & Knowledge Graphs
Details
🚀 The End of YAML? Managing Infrastructure with Natural Language & Knowledge Graphs
What if you could manage your infrastructure just by describing what you want in plain English?
❌ No Terraform files.
❌ No Bicep modules.
❌ No YAML.
Just natural language — and an AI system that understands your intent, reasons about your environment, and produces the exact operational actions required.
This isn’t a chatbot guessing commands. It’s a new paradigm powered by RDF/OWL knowledge graphs, giving structure and meaning to your infrastructure so AI can interpret, validate, and safely execute requests.
⚡ Live Demo Included!
Watch infrastructure get created, inspected, and fixed — using only natural language instructions.
👀 What We’ll Explore
See how everyday requests translate into precise, validated operations:
🗣️ “Spin up a secure web server with auto-scaling.”
→ Generates real CLI/API operations.
🔍 “Show me everything that violated compliance last week.”
→ Instant answers from the knowledge graph.
⚡ “Fix the public-IP violation on the production VM.”
→ AI detects drift and proposes remediation.
🔄 “What breaks if I delete this security group?”
→ Natural-language impact analysis.
☁️ “Deploy a three-tier application across environments.”
→ Describe intent — AI handles orchestration end-to-end.
🧠 How It Works (Deep Dive)
We’ll unpack the architecture behind this emerging model:
• Define infrastructure concepts, policies, relationships, and events with RDF/OWL ontologies.
• Store resources, versions, dependencies, and compliance rules in a Knowledge Graph (e.g., GraphDB).
• Convert natural language to structured intent and safe operations using LLMs.
• Apply validated changes across cloud and Kubernetes with CLI/API executors.
This is a deterministic system with AI as the interface — not AI improvising infrastructure.
🔥 Addressing the Hard Questions (Open Discussion)
• Can we trust natural language for mission-critical changes?
• How do we prevent unsafe AI-generated actions?
• Is this simpler than template-based IaC, or just different?
• What’s mature today and what’s still evolving?
We’ll approach these honestly — no hype, no hand-waving.
👨🏫 Speaker
Sacha Lin, Devops Engineer, Xdevops.ai
Specializing in AI-driven DevOps, semantic modeling, and next-generation automation systems.
🎯 Perfect For
• DevOps, SRE, and Platform Engineers
• Architects exploring next-gen cloud automation
• Anyone curious about knowledge graphs, RDF/OWL, or LLM-driven operations
• Engineers who’ve ever thought: “There must be a better way than YAML…”
Have experience with AI assistants, IaC pain points, or strong opinions about natural-language interfaces?
We’d love to hear from you — community speakers are always welcome!