AI Meets IaC — Intelligent Terraform Authoring on Azure
summary
AI generates runnable Terraform for Azure by combining GitHub Copilot in VS Code, a new Azure Terraform MCP server, and a portal full-screen Copilot editor that shows tree view, green/red diffs and a draft configuration with Copilot flow. The system performs deterministic TF validation and retrieval grounding, auto-installs VS Code extensions and MCPs, and assembles stacks from AVM AZ API and Azure RM modules (demo builds a stack in ~15 minutes sped up). Planned upgrades include GPT-5 model and a drift tool called Absorb.
AI-written summary · about
chapters (8)
- 0:00Welcome
- 1:15AI and Terraform focus
- 2:36Azure Copilot Terraform
- 10:57MCP server for Terraform
- 16:16AI-driven Terraform demo
- 20:30Future AI plans
- 26:56Q&A on CICD pipelines
- 30:48Closing
related talks
25mAccelerating Azure IaC: Copilot, Terraform, and the power of AI
48mState of Infra-as-Code
50mInfrastructure as Code for Azure with Terraform
35mWhat’s new in Terraform for Azure?
37mTerraform & CI/CD
31mMaking AI work for you: Terraform engineer blueprint
06mBoosting developer productivity with an AI-powered developer portal
01h 05mAn Infrastructure as Code Bake-off, comparing ARM, Terraform and Bicep