I started my career as a graphic designer. Photoshop, pixels, layers — the whole thing. When Figma & Canva came along, the transition was surprisingly smooth. The primitives were all still there: shapes, typography, layers, components, etc. The tools changed; the mental models mostly didn't. I adapted fast.

Around the same time, I was building sites with WordPress. Then I started learning JavaScript — Vue, specifically. Again, the primitives carried over. I already understood HTML and CSS. JavaScript was the missing piece, and once it clicked, I was frontend-engineer ready.

And that's when things got interesting.

Going frontend-first was a superpower.

As a frontend engineer, I didn't just bring code skills to the table. I brought design thinking, an eye for detail, the ability to take ownership from concept to screen. I had more control than ever. It was awesome.

But the more I built, the more I wanted to understand the full picture.

What was happening on the other side of the API call? How were the data models structured? Why did some systems feel solid and others brittle? The curiosity kept pulling me in, and at some point I thought: I could learn this too.

And so I did.

It helped that the community was already moving in this direction. JAMstack, Nuxt.js, Next.js — frameworks that brought backend primitives to the frontend. Server-side rendering, API routes, middleware.

The pattern was familiar: new tools, transferable mental models, adapt and ship.

AI hasn’t changed the pattern. It has accelerated it.

Here's what AI actually does for engineers right now: it collapses the time it takes to internalize unfamiliar systems.

The parts of backend development that used to feel uncomfortable — the request/response layer, CRUD operations, database modeling, service integrations, queues, background jobs — can now be scaffolded, read, broken, and understood in days. Not just because AI does the work for you, but because it gives you a patient, always-available collaborator to think through systems with.

I'm now genuinely full stack.

Not "full stack" as a resume word — full stack as in: I can own a feature from database schema to UI animation and make confident decisions at every layer.

You can be too.

Here’s what this means for you:

The builders who are going to thrive in the next few years aren't the ones who hand off to AI. They're the ones who use AI to expand their surface area — to understand more of the stack, own more of the system, and ship with fewer dependencies.

This means:

• Learn aggressively: Use AI to explore systems you've been intimidated by. Ask it to explain a message queue, walk through a database index, scaffold an auth system. Then break it and rebuild it until you actually understand it.

• Understand systems, not just syntax: The goal isn't to write perfect Go or Rust. The goal is to understand how pieces talk to each other — and AI makes that accessible faster than ever.

• Ship hard: The barrier to building real things has dropped significantly. The builders who take advantage of that window will separate themselves.

The through-line of my career has been this: every time a new tool or paradigm showed up, the primitives I'd built transferred over. The tool changed but the thinking compounded.

AI is the same. The designer, engineers, and frankly anyone, who treats it that way — as an accelerant on top of real understanding, not a replacement for it — are the ones who will build the most meaningful things in this shift.

So re-leverage. Understand systems. And ship, fast.