We build infrastructure that survives the cloud decision being wrong. Terraform-first, environment parity from local to production, and pipelines that do not care whether the underlying provider is AWS, Azure, or GCP.
What the work involves
Infrastructure-as-code
Terraform modules with sane defaults and tagging
Environment parity: local, ephemeral preview, staging, production
State management, remote backends, and policy-as-code (OPA / Sentinel)
Runtime & pipelines
AWS, Azure, GCP — chosen against the customer, not the engineer
Kubernetes when it pays for itself, serverless when it doesn’t
GitHub Actions / GitLab CI / Azure DevOps — same shape across providers
Ephemeral preview environments per PR for review and design partners
Operations, cost & resilience
Logs, metrics, traces — OpenTelemetry-first, with SLOs and runbooks