People often think that breaking into tech just means learning a programming language and starting to write code. The moment I put this site’s infrastructure online, I understood one thing right away: code is just the final tool. The real challenge is architecture.

To structure my roadmap toward becoming a Cloud Architect, I decided not to rely on chance or standard tutorials, but to talk directly with professionals already established in the field: cloud engineers (working on AWS and GCP) and Machine Learning researchers. Two fundamental pillars emerged from those conversations, and they’re now guiding my studies.

1. System Design Before Servers

The first big piece of advice I got completely upended my plans: I thought I had to start directly with the Cloud, but I was completely wrong.

There’s no point diving into the depths of AWS or Google Cloud if you haven’t mastered System Design first. The various providers are conceptually identical; what actually makes the difference is the solidity of the application you run on top of them.

Before touching infrastructure, my attention is focused on:

  • Design Patterns: understanding proven architectures for solving common software problems.
  • Application architectures: studying in depth the pros and cons of modular monoliths versus microservices or micro-frontends.
  • Orchestration: understanding how batch systems and various services communicate with each other.

Only once the logical architecture is clear does it make sense to open a Cloud provider’s console to pick the right service for a real-world context.

2. Abstracting Programming in the AI Era

The second pillar is about our role as developers in the age of Artificial Intelligence. Today, models like Claude or ChatGPT can write code incredibly efficiently. My goal isn’t to compete with them on syntax, but to generalize and abstract.

People doing high-level research confirmed to me that the key isn’t memorizing a language, but strengthening logic and mathematical foundations:

  • Applied mathematics: linear algebra, calculus, probability, and statistics (using platforms like Khan Academy) are the tools for understanding why an algorithm works, not just how to write it.
  • Continuous experimentation: don’t stop at theory. Dive in, write a thousand different pieces of code, run tests. Learning isn’t linear: at first the concepts seem disconnected, but as you combine practice with theory, the pieces eventually click into place on their own.