← Writing

How I built AWS infrastructure on Free Tier without breaking things

Most tutorials on AWS either skip real-world practices or assume you’re okay spending money.

I wasn’t.

I wanted to build something that felt like real infrastructure — using Terraform, CI/CD, and proper workflows — but still stay within the Free Tier.

That turned out to be harder than expected.

The problem

Working with AWS is easy until it isn’t.

A small mistake in configuration, an extra resource, or a forgotten service can quietly start costing money. Most beginner setups don’t account for this.

At the same time, many tutorials simplify things too much — skipping Infrastructure as Code, ignoring CI/CD, and focusing only on getting something running.

I wanted both:

  • real practices
  • zero or minimal cost

What I tried

Initially, I followed standard setups:

  • manually creating resources
  • experimenting with different services
  • testing things directly in the console

This worked, but it quickly became messy:

  • no reproducibility
  • no clear structure
  • hard to track changes

So I switched to Terraform.

What actually worked

Using Terraform changed everything.

Instead of manually managing resources, I could define infrastructure as code and apply changes consistently.

Then I added a CI/CD pipeline to automate deployments.

The setup became:

  • Code → Git push
  • CI/CD → Terraform apply
  • AWS → infrastructure updated

This made the system:

  • reproducible
  • version-controlled
  • easier to reason about

The real challenge

The hardest part wasn’t writing Terraform.

It was staying within the Free Tier.

I had to:

  • carefully choose services (EC2 over managed alternatives)
  • monitor usage constantly
  • avoid accidental resource creation
  • understand IAM permissions properly

This forced me to think more like an engineer, not just someone following a tutorial.

Key lessons

A few things that mattered more than expected:

  • Cost constraints improve design decisions
  • Infrastructure as Code should be used from the start
  • CI/CD is not an “advanced feature” — it’s a baseline
  • Understanding how resources behave is more important than using many services

Conclusion

This project became less about saving money, and more about learning how to think in terms of systems.

This project started as a way to avoid spending money, but it turned into something more valuable.

It helped me understand how infrastructure actually works — not just how to spin things up, but how to manage, structure, and reason about systems.

I’m continuing to build on this with better automation, monitoring, and more structured environments.