Skip to content

Infrastructure as Code

Terraform architecture, modules, state strategy, brownfield import, and a testing and promotion workflow that stops drift.

The problem

Infrastructure was clicked into existence and later partly described in Terraform that nobody fully trusts. State lives in the wrong place, modules are copies of copies, and applying to production is a nervous manual event.

What you receive

  • Repository and module structure with ownership and versioning rules
  • Existing infrastructure under management, with drift reported rather than discovered
  • A promotion workflow from plan to production with the right approvals
  • Documentation the next engineer can onboard from

Capabilities

  • Terraform architecture and repository design
  • Reusable, versioned modules with tests
  • State strategy, locking, workspaces and environment layout
  • Brownfield adoption: importing and reconciling existing infrastructure
  • Policy enforcement (OPA, Sentinel, Azure Policy, SCPs) in the plan/apply path
  • Environment promotion and drift detection
  • CI/CD integration with plan review and controlled apply
  • Multi-account, multi-subscription, multi-region and tenant-isolation patterns

Example engagement

Bringing a hand-built AWS estate under Terraform

Inventoried the estate, designed the module and account layout, imported resources in a controlled order with plan diffs reviewed at each step, and wired plan/apply into CI with policy checks so that production changes stopped being a person at a laptop.

Representative of the work; client details are not published.

Technologies

  • Terraform
  • OpenTofu
  • Terragrunt
  • OPA / Conftest
  • Sentinel
  • Terratest
  • Azure Policy
  • AWS SCPs

Before you write to us

The questions we get asked first.

Answered the way we would on a call. The consulting FAQ covers how engagements work in general.

Our estate was built by hand. Can it be brought under Terraform?
Yes, incrementally. The order that works is import what exists into state with no changes, then refactor into modules, then let the pipeline own applies. A big-bang rewrite is how brownfield adoptions fail; the guide on this site describes the staged route.
Terraform, OpenTofu, Bicep or Pulumi?
Whichever your team will maintain. The decisions that outlast the tool are state and locking, module boundaries, the pipeline identity that applies, and how an exception is recorded. We have shipped all of the above and recommend based on the estate, not on preference.
What access do you need?
Read access to the infrastructure code, the state backend, and the cloud accounts it manages, and a few hours of the team's time. Nothing is run under your production credentials from outside your own tooling, and write access only arrives with an implementation phase, through your pipeline and your review.
Who does the work?
Senior engineers who have run this kind of estate, working in your repositories, pipelines and rituals. There is no hand-off from a sales team to a delivery team: the people on the first call are the people on the engagement.