Design in the age of AI
AI isn't replacing my design process, but it's removing friction inside it, so I can spend more time on the decisions that actually matter:
-Structuring messy thinking. Interview notes, rough flow ideas, an unclear case study section: I use AI to turn scattered thoughts into a first draft I can react to, not to generate ideas for me.
-Speeding up repetitive work. Sprint tickets, research questions, testing summaries, adapting case study sections: AI handles the first pass for me, so I can focus on refining, not starting from zero.
-De-risking prototypes. On my CRM project, I used AI to quickly mock up interaction ideas: like a playful confirmation animation: before committing dev time to build them. Same logic as a paper prototype, just faster.
-Where I draw the line. AI doesn't decide what users need, doesn't replace real testing, and doesn't get final say on whether something feels right. That judgment is still the actual job.
This way, AI hasn't changed what I design: just how much time I get to spend doing it.
About this portfolio
I cannot count how many times I've rebuilt this portfolio. I had to say that I didn't land on my portfolio, I stumbled into it after trying (and abandoning) almost every tool a designer could reach for:
-Behance: felt like moving into a house someone else already decorated. Cute, but not mine, I need more interactions, more control.
-PDF: basically emailing people a brochure and hoping they'd imagine the interactions themselves.
-Notion: i'm totally in love with this tool, but not to create a portfolio, it looked like a very organized to-do list.
-Framer: so close! Right up until I wanted one specific animation and the tool looked at me like I'd asked for too much.
-Wix: fast, easy, done in a day, also the exact ceiling I hit the moment I wanted anything custom and personal.
Then, I just wrote the code myself, and suddenly nothing was standing between me and the thing in my head. Every animation, every weird little detail: mine, on purpose,and no platform telling me "no."
So if this site is live and running right now, just know it's held together by an unreasonable number of late nights, but it totally worth it!
What NOT to design
Good design is knowing what to leave out, and why.
-CRM project: the agency wanted a dashboard with dozens of widgets. But the real problem was too many steps to publish a listing, not too few options — so I cut it to the three things agents actually checked daily. Fewer features, better time-to-publish.
-EdTech project: AI could've been the headline feature. Instead I made it invisible infrastructure — adapting onboarding quietly instead of asking users to configure it.
-The pattern: adding more usually proves effort, not value. My job is finding what actually solves the problem, and defending cutting the rest — even when a stakeholder loves the shiny version.
That's the part that never shows up in a Figma file.