Skip to main content

FAQ

Here are some questions no one asked yet.

What's your Spring Boot background?

I've been working with Java and Spring Boot professionally since late 2022. My two most relevant roles:

  • Applica Solutions (Nov 2022 – Apr 2023): Standardized the backend REST APIs for Argentina's first QR-based payment system.
  • Devsu (May 2023 – Oct 2025): Integrated identity providers with customer-facing pages and applications for Ecuador's largest private bank, working with Google Cloud and Microsoft Azure.

Before any of that, I took a Spring Framework fundamentals course at UTN Buenos Aires, which got me started. For the full timeline, see About the author.

Will I come across codebases similar to the one explained by these docs?

Not really. Real-life codebases are messy. They have years of accumulated decisions, workarounds, half-finished refactors, and comments nobody remembers writing.

The codebase alongside these documents is intentionally kept clean and straightforward. That's an educational choice — the goal is to show how things should work without the noise, so you can recognize the patterns when you encounter them in the wild.

Do I really need to know this much Spring Boot to get things done?

Kinda, but not really.

It took around 20 documents to get just a working GET endpoint. That sounds like a lot, and it is. But the point is taking the time to get the foundations right, so that adding features later becomes trivial. If you skip the groundwork, you'll eventually hit a wall where every new feature requires fighting the framework instead of using it.

Think of it as paying upfront: invest now, save later.

Are these docs and/or code examples AI generated?

Kinda yeah. The process roughly goes like this:

  1. I build a new feature in the gradle monorepo (I use AI here too — at the time of writing I don't have a predefined workflow, it's mostly planning, prompting, and manually reviewing). Then I push it.
  2. I run the doc-from-commit opencode command I created, which generates a new Docusaurus doc page from a topic and GitHub commit link(s).
  3. I manually review what the AI produced. It's usually rough around the edges, but it gives a solid first version.
  4. I run the review-doc opencode command I created, which reviews a doc page and its TSX companion for style and compliance issues (good for double-checking).
  5. I do the Spanish version using a custom translate-doc opencode command. Then I manually check for weird translated phrases — I'm a native Spanish speaker, so a quick read is enough to catch the odd ones.
  6. I push. This project has a GitHub Action that does a quick build check, then notifies a Coolify instance for deployment.

So yeah, AI is involved at multiple steps, but everything gets human review before it ships.