UXnicorp Academy

Building a modern web development course with Astro

"A performance-oriented educational platform that demonstrates, through its own architecture, the principles of speed, simplicity, and Zero-JS that it teaches."

Career

FIG 01. Structural representation of the solution© Gonzalo Daniel Vega

01 / CONTEXT

The starting point of the challenge

Most web development courses in Spanish shared several problems: outdated content, too much passive theory, poor performance, and extremely heavy platforms loaded with JavaScript that penalized Core Web Vitals. At UXnicorp, they needed a consistent way to onboard developers, document their working stack, and teach their product-centered design methodology. I decided to build a top-quality educational resource — designed for the team but open to anyone who wanted to learn. No logins, no payments, no barriers. If it helped someone else too, all the better.

02 / RESEARCH

Digging beneath the surface: What did I investigate?

I analyzed platforms like freeCodeCamp, Udemy, Coursera, MDN, and YouTube tutorials. I observed four main problems:

Too much passive video, which reduces retention and makes quick reference to concepts difficult.
Ignored performance: educational platforms that used heavy SPAs while, paradoxically, teaching good frontend practices.
Static or non-editable examples, drastically separating theory from practice.
Early abstraction: courses that taught React before explaining HTML, the DOM, or how a browser actually works.

I understood that the course had to feel fast, clear, progressive, and interactive. And technically, the product had to be a reflection of what it taught: if you teach performance, the website has to load instantly.

INSIGHT REVEALED

THE SYNTHESIS

"I defined key principles: Zero-JS where it's not needed (without loading heavy SPAs for static text), progressive interactivity (React only where it adds value: playgrounds, demos, theme toggle), content-first, and non-linear learning without unnecessary backend. The project made something evident that I had already been thinking about: architecture also communicates criteria. You can't teach good performance practices on a slow website."

Understanding this pillar completely transformed the tactical direction of the project, allowing unnecessary noise to be filtered out and focusing on strict value.

03 / TACTICS

Tradeoff Analysis: options on the table

Building custom software requires honestly evaluating scenarios. No architecture is perfect — every decision has a price:

OPTION 1

Next.js (App Router)

Advantage: excellent ecosystem and mature SSR. Disadvantage: too much overhead and unnecessary JavaScript for a static content site.

OPTION 2

Gatsby

Advantage: content-oriented. Disadvantage: less modern DX and extremely slow build times.

OPTION 3

Astro (Islands Architecture)

Advantage: renders static HTML, eliminates unnecessary JS, excellent performance, and perfect for educational content. Disadvantage: newer ecosystem and certain limitations with complex interactivity.

04 / DETERMINATION

The Decision: The path chosen

I chose Astro as the core of the project. This allowed me to create ultra-fast static content and add encapsulated interactivity through "React Islands" in playgrounds and interactive components like the theme toggle or code viewer. To avoid an unnecessary backend, I kept the site fully static with Astro in SSG mode, without databases or external APIs. The decision not to have auth or a backend wasn't just technical: I wanted anyone to be able to enter and learn without friction. Search, progress tracking, and authentication were left out of the initial scope to prioritize deployment speed and simplicity. Later on, I migrated the content to MDX + Content Collections, separating content from structure and making the project more maintainable.

WHAT ACTUALLY WORKED

Astro + Islands was the right call. The site is blazing fast, the interactive demos add value without making it heavy, and having no auth or backend doesn't just simplify everything — it leaves it open for anyone who wants to learn.

LIMITATIONS & TRADE-OFFS

Without a backend, progress tracking across devices is lost.

FROM GONZALO'S PERSONAL LOGBOOK: WHAT WOULD I DO DIFFERENTLY TODAY?

"I'd like to add more interactive demos in other modules, not just in CSS."

Retrospective note after launch

Gonzalo Daniel Vega — 2025 — SFV Catamarca, Catamarca