Scalr

Scalr is an infrastructure as code platform for Terraform and OpenTofu that gives platform teams remote runs, policy checks and drift detection, billed per run.

Scalr replaces Terraform Cloud as the shared place where an organisation plans, approves and applies its cloud changes. It keeps state, variables, policies and run history for every workspace in one account, so each change follows the same review path. Each workspace runs Terraform or OpenTofu, and Terragrunt projects are supported too. Scalr, Inc. develops the platform and sells it as the company's one product. A run starts with a plan, then waits in an approval phase unless the workspace has auto apply turned on. Opening a pull request in a linked repository triggers a dry run, and the result is posted back as a comment. Open Policy Agent policies check each plan at advisory, soft or hard enforcement, and Infracost adds an estimate of the cost change. Drift Detection compares live resources with the code on a schedule per environment, and the Run Scheduler can destroy development workspaces on a timetable. The Private Module Registry holds versioned modules, and agent pools execute runs inside a company's own network. Platform engineering teams that look after many workspaces for several application teams are the people it is built for. IAM access policies at account and workspace level let them hand out self-service access while keeping apply rights narrow. Companies with resources in private networks keep execution inside those networks through self-hosted agents. Scalr is bought by usage, counted in runs rather than users or managed resources, and larger organisations arrange an Enterprise agreement through sales. Most work begins with a commit: a pull request produces a plan, reviewers read it beside the policy results, and an approver confirms the apply. That confirmation can come from the Scalr interface, the API, Slack or Microsoft Teams. Provider configurations connect workspaces to AWS, Azure and Google Cloud, and sensitive variables stay masked in plan and state output. Federated environments let one workspace read the outputs of another, so dependent stacks stay wired together without copied values. For a team leaving Terraform Cloud, it is a way to keep running the same Terraform code under central guardrails.