About the role
Why the best SREs write code like devs, not sysadmins
You want to break into cloud engineering, platform, SRE, or DevOps?
Cool. Learn to code.
I don't mean hacking together some Terraform modules you found on GitHub. I mean writing real code. Scripts, daemons, tests, CLI tools. Treating infrastructure like a software product. Because the truth is: the best infra engineers think like developers.
๐ง From Dev to Infra โ Why It Matters
When I look around at the strongest platform engineers I've worked with โ folks building production infrastructure at scale โ they all have one thing in common:
They were developers first.
Or they learned to think like one.
They write clean, tested, debuggable code.
They design systems, not scripts.
They reason about performance, observability, dependency graphs, build pipelines, and rollback strategies.
They don't just automate.
They engineer.
๐งต Infra Without Code Is Fragile Theater
Sure, you can "get by" with some CLI kung fu and a dozen copy-pasted YAMLs. But if you're managing systems that people depend on โ customer-facing services, payment platforms, large-scale pipelines โ that duct tape won't cut it.
You'll be asked to:
Debug distributed systems in production
Write operators, daemons, or lifecycle controllers
Build reliable CI/CD platforms
Orchestrate ephemeral environments
Implement backpressure, rate limiting, retries
None of that is point-and-click.
It's software.
๐ฏ Trying to Break In? Focus on This First
If you're aiming to land a cloud/SRE role, here's your roadmap:
Learn a real language (Python, Go, or Rust if you're spicy)
Build a CLI tool for your dotfiles or homelab
Automate infra deployments with CI/CD
Write tests for your infra-as-code (TDD with OpenTofu? Hell yes.)
Contribute to open source infra tooling
Read the source code behind tools like kubectl, helm, gh, colima, or cosign
That's how you think like an engineer.
Not a technician.
Not a sysadmin.
A builder.