I take ambiguous B2B SaaS problems to shipped product, and build the systems that make the next one faster.

Currently at Frequency, leading design for Monetize and building the design system and AI harness behind it.

Role
Systems UX Designer, Frequency
Focus
B2B SaaS, design systems, AI workflows
Based in
Los Angeles

Selected work

Before-and-after comparison of AI-generated Frequency Studio screens: a rough mid-June version with placeholder labels next to a current version that closely matches the production Manage videos list view.

Frequency Studio · Nov 2025 to present

From style guide to LLM-ready design system

A style guide of suggestions left screens drifting apart and nothing an AI could reliably build from. I turned it into 485 tokens and 104 written component rules, then built the harness that generates wireframes from it.

Role
Systems UX Designer, built solo
A call I owned
I chose a three-tier token architecture over the faster two-tier, accepting a slower start rather than leaving us something to unpick as the product grew.
Screens from the Monetize module in Frequency Studio: the distributor dashboard, an edit-endpoint panel with supply ad tags, a placement-rule configuration with ad format toggles, and the ad creatives list.

Frequency Studio · 2026 · In development

Monetize: designing a 0→1 in-scene advertising tool

Frequency opened a new revenue avenue: ads rendered inside the video. It needed a workflow for placing those ads and monitoring what they earned, and none of it existed. I designed the core screens against a conference deadline.

Role
Lead Product Designer
A call I owned
I explored a requested bulk-review feature dozens of screens deep, found no user need it served, and worked through its purpose with the PM and the CEO until we agreed to drop it.

Earlier work: Oracle projects, research, and UX teaching

How I work

Every decision traces to a user need.

If I can't name the need a feature serves, I ask why until someone can. When no one can, I bring the evidence to the team and work to persuade stakeholders that our time is better spent elsewhere. Descoping is a design decision.

Constraints are inputs, not obstacles.

Deadlines, legacy systems, and business goals that don't align neatly with user goals are part of the problem statement. Good design is finding what serves the user inside what's real.

AI accelerates the work; judgment stays human.

I use LLMs to explore faster and check my blind spots. A generated screen is a starting point, and I evaluate and tweak the designs to make sure they address a verified user need.

About

I studied cognitive science with a computer science specialization at UCLA, started my career in cloud engineering at Oracle, and switched into UX from the inside, which means I design with the people who have to build it in mind.

More about me →