This is a submission for the New Year, New You Portfolio Challenge Presented by Google AI
About Me
I'm a backend-focused software engineer working in the banking domain, with a growing interest in AI-assisted development and how modern tools can reshape the way we build software.
This portfolio wasn’t planned in detail or designed with a big upfront architecture. Instead, it came from an experiment I kept simple on purpose:
What happens if I vibe-code a portfolio using only two prompts?
Portfolio
How I Built It
The entire portfolio came from two prompts.
Prompt 1 — Initialization (Antigravity IDE)
I started inside Antigravity, a Google IDE with integrated AI agents.
I didn’t provide wireframes, assets, or detailed specs — just a loose idea:
“I want a portfolio that feels like a 2D pixel-art game start screen, showing my stats, skills, and achievements.”
The AI agent suggested a concrete plan:
- a game-style UI
- character stats as skills
- inventory as tech stack
- achievements as milestones
All I really did was approve the plan and iterate lightly.
That alone got me to a working, polished portfolio that already felt good.
Prompt 2 — Making It “Wow”
At that point, the site worked and looked fine — but it felt like a game start screen without an actual game.
So I asked a second, more intentional prompt:
“This is good, but not great. How can I turn this into an experience, not just a screen?”
The response reframed everything:
- sections became locations
- content became quests
- navigation became exploration
- progression replaced scrolling
I fed that prompt back into the tool and refined the experience with:
- light state management
- dialogue-style text
- small progression cues
No new assets. No heavy animations. Just better structure and intent.
What I'm Most Proud Of
What I like most isn’t the UI — it’s the process.
With just:
- one prompt to generate structure
- one prompt to add depth
I ended up with a portfolio that feels playable, personal, and memorable.
It changed how I think about building:
- less over-planning
- more direction-setting
- letting AI handle the “how” while I focus on the “why”
This project convinced me that vibe-coding, when guided properly, isn’t lazy — it’s strategic.
Top comments (0)