⬆︎

web designer

and builder

I design and ship in Framer: no handoff, no translation loss. Eight years of product thinking behind every page.


From cold template to trusted coach

Context

Mental health coach with a warm Instagram presence and a cold, generic Wix website.

Problem

Potential clients were losing trust at the first touchpoint, the website didn't feel like Gabriel.

Constraints

The same organic typography and hand-drawn style every other coach was using.

Key Decision

Treated every section of the website as a conversation, not a page to fill.

Insight

If your social presence is warm and your website is cold, you've already lost the client.

Impact

+25% session duration one month after launch.


A Design System for legacy tools

Context

Solo designer at an org where the team had never used a design system.

Problem

INTS had brand inconsistencies across touchpoints.

Constraints

Everything had to work in Divi, the tool they already used on their website.

Key Decision

Simplify components so anyone could build them.

Insight

I sold the system to leadership.
I should have sold it to the people using it.

Impact

~50 components built, pages shipped, style guide live but never fully adopted.


Engineering the tech job search

Context

Reading AI Snake Oil sparked a question: could probabilistic thinking actually fix the job search?

Problem

Candidates don't know where they're failing.
They apply more, burn out faster.

Constraints

No team, no budget.
I had one day to build something real.

Key Decision

Build a forecasting engine to get one number: how many applications you need.

Insight

Show people their own data, and the anxiety drops.
A hopeless process becomes a solvable puzzle.

Impact

First test exposed two unresolved problems: unclear onboarding and a number that felt crushing.


Validating Vehicle Concepts with VR

Context

Ford needed to test early-stage vehicle concepts before building physical prototypes.

Problem

Translating physical HMI to VR while maintaining strict realism for valid behavioral testing.

Constraints

Compromising the realistic feel of the vehicle interior just to make digital interactions easier.

Key Decision

Prioritize realism to accurately test driver behaviors.

Insight

Any unnatural interaction in the virtual environment would skew behavioral data.

Impact

Established a repeatable VR testing protocol.
Won 2nd place at the IEL Careers Award 2022.


Can you trust an AI to know your taste?

Context

Agency simulation for a $60M food tech brand.

Problem

Create social content that converts without triggering the "this feels like an ad" response.

Constraints

AI tooling — no studio, no photoshoot, two weeks.

Key Decision

Reject AI's default instincts and apply an editorial layer.

Insight

Hungryroot doesn't sell food. It sells the removal of a specific kind of exhaustion. Every asset had to speak to that.

Impact

A multi-channel asset suite across Stories, Reels, and Hero Banners — built with a replicable AI-to-editorial workflow.


Eight years of product thinking behind every page

Over 8 years working across health, mobility, and games taught me that the best interfaces are the ones nobody notices. I bring that same thinking to the web: clear hierarchy, intentional motion, and pages that load fast and convert clearly. I design and build in the same tool, which means what you see in the mockup is what goes live.