AI to Build My Portfolio | Here’s What It Got {Right

The Experiment

Earlier this year, I decided to revamp my personal portfolio. It was outdated, inconsistent, and frankly, a little embarrassing. But between work and life, the idea of coding a new site from scratch felt like climbing Everest in flip-flops.

So I had a thought: What if I let AI help me?

I’d been playing around with ChatGPT and GitHub Copilot, and I was curious. Could artificial intelligence actually build something as personal and nuanced as a portfolio site?

To test that, I made a rule for myself: I’d let AI guide every major step — from ideation to code to copy.

Here’s how it went down — and where it nailed it, and where it fell on its (virtual) face.

LEARN MORE : https://digitalmohit.co.in/

Step 1: Brainstorming the Structure

I prompted ChatGPT:

“Create a sitemap for a modern personal portfolio for a software developer.”

Within seconds, it gave me this:

Home
About Me
Projects
Blog
Resume
Contact

It also suggested features like a dark/light theme toggle, smooth scrolling, GitHub integration, and a tech stack section with icons. Pretty standard stuff, but solid. I asked it to explain why it chose each section — and the reasoning made sense.

What AI got right:

Fast, sensible layout suggestions
Consideration of UX trends (responsive design, accessibility, etc.)

What it got wrong:

Everything felt a bit generic
No real personality — which is kind of the whole point of a personal site

Step 2: Writing the Copy

This part impressed me. I asked ChatGPT to:

“Write a friendly, professional About Me section for a full-stack developer who loves React, design systems, and coffee.”

Here’s a snippet of what it wrote:

“Hi, I’m Alex — a full-stack developer with a passion for building intuitive, scalable web experiences. When I’m not writing clean code, I’m probably designing component libraries or experimenting with new UI frameworks — usually with a cup of coffee in hand.”

It wasn’t perfect, but it felt… usable. Like something I could tweak and make mine.

What AI got right:

A great first draft for copy
Tone was surprisingly natural
Saved me hours of staring at a blinking cursor

What it got wrong:

Clichés (“clean code,” “scalable experiences”)
No real voice — it needed editing to sound like me

Step 3: Design & Aesthetics

I tried asking DALL·E and other image tools to generate a custom background or hero image. Some of the outputs were amazing. Others looked like rejected album covers from a 2006 synth-pop band.

For UI, ChatGPT was great at suggesting color palettes and font pairings. But again — execution mattered. AI might suggest “use Inter and #1f2937,” but only I could feel when it looked right.

What AI got right:

Helped me make faster visual decisions
Suggested color and font pairings that actually worked well

What it got wrong:

No emotional sense of brand
Didn’t understand my audience or personality

The Unexpected Benefits

One thing I didn’t expect? Using AI forced me to explain my ideas clearly — in writing. Every time I prompted it, I had to reflect on what I actually wanted. That act alone helped me clarify my goals, aesthetic, and message.

Also, I realized I didn’t hate writing code — I hated starting code. Once AI gave me a base, I was excited to jump in and make it better.

So, Can AI Build Your Portfolio?

Yes — but only up to a point.

AI is a fantastic creative partner. It’s like hiring a really fast (but slightly clueless) intern. You’ll still need to:

Curate the content
Infuse personality
Refactor code
Make design decisions
Bring the human element

But if your goal is to avoid the inertia of starting? AI’s your best friend.

Final Thoughts

Building my portfolio with AI wasn’t frictionless — but it was way more fun than doing it solo. I got to spend more time being creative and less time fighting with CSS.

Would I do it again? Absolutely.

But next time, I might just prompt: