Infrastructure as code

Infrastructure as code means describing networks, databases, clusters, permissions and DNS in files that live in version control, and letting a tool work out and apply the changes. It is how most platform teams keep environments consistent and reviewable. The tools here either define that code, run it as a team, or add AI help to write and review it.

Terraform is the most widely used language and tool, and OpenTofu is its open-source fork run by the Linux Foundation after Terraform moved to the Business Source License. Pulumi takes a different route and lets you write infrastructure in TypeScript, Python, Go and other general-purpose languages. AWS CloudFormation is the native option for teams that live entirely on AWS, and Ansible, with Red Hat's Lightspeed assistant, covers configuration of the machines themselves.

Platforms such as Spacelift and env0 sit on top of these tools. They run plans and applies in a controlled place, keep state safe, ask for approval before risky changes, show cost estimates and spot drift between code and reality. Firefly starts from the other end and finds resources that were created by hand so they can be brought under code. Brainboard lets teams draw an architecture and get Terraform back.

Featured AI DevOps tools: infrastructure as code

SponsoredInfrastructure as code This position is open Put your tool at the top of the infrastructure as code page. Paid spots are always labeled Sponsored. From $99 a month. Get featured

What a featured listing costs ยท from $99 a month

9 AI DevOps tools in infrastructure as code

OpenTofu

by Linux Foundation

Open-source fork of Terraform under the Linux Foundation, compatible with most Terraform configurations and providers.

  • Free plan or free tier
  • Open source
  • Works with AWS

Pulumi

Infrastructure as code in general-purpose languages such as TypeScript, Python and Go, with an AI assistant in Pulumi Cloud.

  • Free plan or free tier
  • Open source
  • Has an API

Spacelift

Platform for running Terraform, OpenTofu, Pulumi and other IaC workflows with policies, approvals and drift detection.

  • Free plan or free tier
  • Works with AWS

AWS CloudFormation

by Amazon Web Services

AWS's native infrastructure as code service, which creates and updates AWS resources from JSON or YAML templates.

  • Free plan or free tier
  • Works with AWS

Terraform

by HashiCorp

Widely used infrastructure as code tool that defines cloud resources in HCL files and applies changes through plan and apply.

  • Works with AWS
  • Works with Kubernetes

env0

Infrastructure as code automation for Terraform, OpenTofu and other tools, with cost estimates, approvals and environment governance.

  • Works with AWS

Firefly

Finds cloud resources that were never written as code and helps turn them into Terraform or other IaC, then tracks drift.

  • Works with AWS

Brainboard

Visual designer that turns cloud architecture diagrams into Terraform code and back, for teams that plan infrastructure visually.

Ansible Lightspeed

by Red Hat

Red Hat's AI assistant for writing Ansible playbooks and tasks from plain-language descriptions.

AI features in this space mostly help with writing and reading: generating a module from a description, explaining a plan in plain language, or suggesting a fix for a failed apply. They do not remove the need for review. Choose tools that keep state and secrets protected, log who changed what, and work with the policy checks you already run.

Whatever tool you pick, a few habits make the biggest difference. Keep state remote, locked and backed up, since it describes your whole estate. Split large configurations into modules with clear owners, so a change to networking does not need the whole platform team to review it. Run plan on every pull request and post the output where reviewers will read it. Add policy checks for the rules you care about, such as no public storage and required tags, so they are enforced the same way every time. And schedule drift detection, because someone will eventually change a setting by hand during an incident, and you want to find out before the next apply quietly reverts it. AI assistants help most when these basics are already in place; without them they just produce more unreviewed changes faster.

Questions about infrastructure as code

Terraform or OpenTofu?

Both use the same language and most providers. OpenTofu is open source under the Linux Foundation; Terraform is under the Business Source License and backed by HashiCorp's commercial products. Check the license against your use and test plans before switching.

Why use a platform such as Spacelift or env0?

They give a team one controlled place to run plans and applies, with approvals, policies, state handling and drift detection. Small teams can manage without one; larger teams usually want the audit trail.

Can AI turn existing cloud resources into code?

Some tools, such as Firefly, find resources created by hand and generate code for them. Expect to tidy the output and to test it with a plan that shows no unexpected changes.

Is Pulumi harder to review than Terraform?

It can be, because general-purpose code allows more abstraction. Teams that agree on simple patterns and keep tests and previews in review find it as manageable as HCL.

Related categories