Clicking around a cloud console isn't a skill anyone hires for. What gets you in is being able to build the same environment twice from code, lock down who can touch it, and explain why the network looks the way it does. Learn it in this order and skip the service catalogue tour.
| Certificate | Best for | Effort | Worth it? |
|---|---|---|---|
| AWS Certified Solutions Architect Associate | People aiming at AWS shops who want the most recognised entry badge on their resume | a couple of months of evenings | Worth it. Recruiters search for it by name, and studying for it fills the gaps in networking and storage that self-taught people tend to have. |
| Microsoft Certified: Azure Administrator Associate | Sysadmins moving from Windows server rooms into Azure | a couple of months of evenings | A strong pick if the jobs near you run on Microsoft. It's hands-on enough that passing it means you can actually do the daily work. |
| Google Associate Cloud Engineer | People targeting teams that run on Google Cloud | several weekends | Good, but narrower demand. Take it when a posting names Google Cloud, not as your first cloud badge. |
| HashiCorp Certified: Terraform Associate | Anyone who already knows one cloud and wants proof of infrastructure as code | a few weekends | Cheap and quick, and a nice pairing with a provider certificate. On its own it won't carry you, since your Terraform repo says more. |
Exam names, versions and prices change often, so check the provider's own certification page before you book.
Worked with AWS and Terraform to manage cloud infrastructure for the company.
Rebuilt 3 hand-made AWS environments as Terraform modules with plan-on-PR in GitHub Actions, cut new environment setup from 2 days to 40 minutes, and trimmed the monthly bill by $8,400 by right-sizing 60 instances.
Yes, if you set a budget alert on day one and tear down what you build with terraform destroy when you're done. Surprise bills usually come from a forgotten NAT gateway or a large instance left running over a weekend.
You need to read and write code comfortably, not build applications. Terraform, YAML pipelines and short Python or Bash scripts cover most days. If you can debug someone else's script and add a feature to it, you're where most teams need you.
Learn the basics, but don't let it hold you up. A lot of cloud teams run on managed container services or plain virtual machines, and you'll pick up the rest of Kubernetes faster once you're watching a real cluster misbehave.
One small app deployed end to end: a VPC, a container service, a managed database, a pipeline that deploys on merge, and an alert that pages you. Add a short README explaining your choices. Depth beats a gallery of tutorial clones.