2026-09-05 –, Europasaal B
I'll explore what it takes to build a fully local AI agent for Terraform code generation, making use of your own private module library, style guides and best practices. I'll also cover realistic model choices for developer hardware and cover how to avoid accidentally turning your laptop into a space heater mid-demo.
If you've read enough terms and conditions to be a little nervous about piping your internal module library and .tfvars files off to a third-party API, this session is for you.
I'll take a look at where local models genuinely help versus where you'll be staring at generated HCL wondering what the HAL (I'm sorry Dave, I'm afraid I can't do that) it was thinking. By the end you'll have a working blueprint for a Terraform pair programmer that respects your air-gap, keeps your compliance team happy, and rides out your WiFi dead zones. And your .tfvars stay yours, we don't talk about what's in them.
Sam McGeown is a Senior Developer Advocate at HashiCorp who spends his time helping developers and platform teams make sense of modern infrastructure. With a background that spans cloud engineering, consulting, and product roles, he has worked across public and private cloud platforms, Kubernetes, APIs, and Infrastructure as Code.
Sam is passionate about turning complex technical ideas into practical, real-world solutions. Whether through live demos, workshops, or conference talks, he enjoys showing how tools can simplify multi-cloud infrastructure and help teams move faster with confidence.