If you’ve tried “AI writes Terraform,” you already know the truth: generating HCL is not the hard part. The hard part is everything that happens after the code exists. Does it match your modules and conventions? Does it widen IAM in subtle ways? Will it replace critical resources? Will
AI agents can reason about infrastructure. But they still fail in one predictable way: they don’t reliably know what’s actually true right now. That’s what cloud graph intelligence provides: a continuously updated model of assets + relationships + change history. And in an agentic world, that model becomes more,
How AI is Transforming the Way Developers Interact with Infrastructure The promise of DevOps was simple: break down silos between development and operations, ship faster, and empower everyone to take ownership of the software delivery lifecycle. But somewhere along the way, that promise got complicated. Instead of liberation, many development
AI can generate Terraform quickly. That’s not the bottleneck anymore. The bottleneck is getting Terraform changes that are safe in a real org: they match your modules and conventions, don’t quietly broaden IAM, don’t trigger surprise replacements, and don’t “pass compliance” while the real environment drifts.
If your platform/DevOps team feels like a ticket queue, you’re paying the most expensive tax in engineering: toil. Toil is the work that keeps production moving today but quietly prevents you from building systems that keep production stable tomorrow. This article explains what toil is, how to measure
Configuration drift is what happens when your actual cloud diverges from what your Infrastructure-as-Code (Terraform, OpenTofu, Pulumi, CloudFormation) says it should be. If you think you “don’t really have drift,” the more likely truth is: you don’t measure it consistently, or you detect it but don’t close
“IaC-first” is the story most teams tell. ClickOps is the reality most teams live. In 100+ conversations with Heads of DevOps and Platform this year, I keep hearing the same line: “We don’t really have drift.” But the best data we do have points the other way. Datadog’s
People often don’t get what deterministic AI means. They think it’s about making LLMs rigid, predictable, or less creative. That’s wrong. It’s about giving them tools and scaffolding, so their output is trustworthy. Think of your own work as a developer. Maybe for hours you can
Most engineering teams start their cloud journey with ClickOps. It’s intuitive: you log in to the console, spin up a VM, add storage, configure a service - all in a few clicks. ClickOps feels simple at first, but over time it creates real problems: * No single source of truth
According to Firefly, 89% of engineering teams claim to use Infrastructure as Code (IaC). But only 6% actually manage 100% of their infrastructure in code. That gap should make every DevOps, platform, and cloud security leader pause. It means the vast majority of companies are operating in “partial IaC” mode:
As AI becomes embedded into more DevOps workflows, a new challenge is surfacing — one most teams aren’t yet addressing: can you trust the code your AI generates? For frontend or application code, “close enough” might be fine. But in cloud infrastructure, a misconfigured IAM policy, a drift-inducing Terraform fix,
If you thought AI wouldn’t impact DevOps as much as it’s transforming software engineering because DevOps is somehow special or more complex—you might already be behind. AI’s impact on Infrastructure as Code (IaC), the standard for scalable and manageable cloud setups, is already well underway. In