cloud-infra-lab: AWS infrastructure as code with Terraform

The problem: a lot of small teams run on one server set up by hand. It works until someone needs a second environment, a rollback, or to know what changed last week. Reproducible infrastructure fixes that, but it has to be cheap to run and easy to tear down.

What I built: Terraform modules for a standard AWS setup, covering VPC, ECS Fargate, RDS, ElastiCache, S3, a load balancer, IAM and CloudWatch alarms. Deployments run from GitHub Actions using short-lived OIDC credentials, so no long-lived AWS keys are stored anywhere. There is a low-cost mode and a one-command teardown. A second track packages the same services as Helm charts for Kubernetes.

What it deploys: two of my other projects, the Slotwise booking backend and the RAG Engine, so the infrastructure is exercised by real applications instead of a hello-world container.

An honest status: this has been validated offline only, with terraform validate and terraform test against a mocked provider, tflint, checkov security scanning, and helm lint. The Kubernetes track ran live on a local kind cluster. Nothing has been applied to a real AWS account yet, so I quote no AWS cost or latency numbers.

This is a personal project. Code is on my GitHub: chaitanya-mhetre, repository cloud-infra-lab.