About me

I'm Lawrence Lee, a Staff Product Designer focused on complex enterprise systems and AI‑assisted decisions.

I design platforms that help teams catch problems earlier, make confident decisions, and move faster when the stakes are high. My work sits inside complex enterprise environments where fragmented workflows, unclear signals, and operational lag create real business cost.

Lawrence Lee
01 — Experience

Where I've done the work

  1. Jun 2026 – Present

    The Carlyle GroupCurrentProduct Design Consultant

    Leading the design system for internal investment tools, and the design-to-code workflow that ships it.

  2. Apr 2024 – Nov 2025

    The Trade DeskStaff Product Designer

    Led DealDesk, turning contract management into an early-warning system for at-risk deals, and built a shared pattern library.

  3. Oct 2021 – Nov 2023

    RokuLead Product Designer, Roku Ads

    Founding designer on Tarmac, Roku Ads' first campaign management system, and redesigned its audience data platform.

  4. Jan 2018 – Sep 2021

    JPMorgan ChaseSenior UX Designer, Wealth Management Money Movement

    Redesigned account transfers and mobile check deposit for wealth management customers, across product, engineering, and compliance.

  5. Before 2018
    • Accenture SongSenior UX Designer
    • DesignitUX Designer & Strategist
    • Appetizer MobileVisual & Interaction Designer
    • TrueParallelJunior Designer
02 — Philosophy

Good enterprise design is less about perfect screens and more about reducing uncertainty.

So teams can spot risk earlier, move faster, and act with confidence. Here's how I work.

01

Start with the stakes

I align on what's at risk, what “success” means, and which constraints are real.

02

Design the decision, not the UI

I map the key decisions users need to make, and the signals they're missing today.

03

Create shared clarity

I bring people into the same mental model through workflows and principles, so we're solving the same problem.

04

Ship a tight loop

I make a small number of high-conviction bets, measure the deltas, and iterate until the system is resilient.

That's how I turn fragmented workflows into systems that surface what matters and enable action before problems compound.

03 — Teaching

Besides designing, I also teach design.

Teaching forces me to articulate first principles clearly. It's made me a stronger designer and partner, especially in guiding people and helping them grow.

I taught UX design as an adjunct instructor at BrainStation, and I mentor designers on ADPList, where I'm in the Top 1% of Super Mentors.

300+designers mentored
3,000+mentorship hours
10juniors into their first design job
Get in touch about teaching
Lawrence teaching a design class at BrainStation
04 — How I work

How I work today

I use AI tools where they save real time, and keep the decisions with people. Two places that shows up in my recent work.

In my workflow

Design to code, with the rules intact

At Carlyle, Figma stays the source of design intent, and a Claude Code skill reads the specs. Engineers move faster without losing the rules, and a typical component went from 2–3 days to about one.

Unified design system
In my product

Signals people can act on

At The Trade Desk, I designed AI-assisted health scoring for DealDesk with human oversight: signals sellers can understand, override, and act on, so at-risk deals surface in 1–2 days instead of weeks.

DealDesk

How I build today

Every project moves through the same four steps. Tools like Claude and Figma handle the legwork inside each one; I set the direction and make the call at every handoff.

MeSets the direction and makes the call at every step
  1. I frame the problem
    01 · Plan
    Research and framingClaude

    Pulls docs, research, and past decisions into one short brief.

    Problem brief
  2. I critique and pick
    02 · Design
    ExplorationFigma · Pencil

    Explores layouts with the real tokens and components.

    Screens in Figma
  3. I review the code
    03 · Build
    Prototype in codeClaude Code · Cursor

    Turns the Figma spec into working code, using the design system's rules.

    Working prototype
  4. I decide what ships
    04 · Feedback
    Review and shareGitHub · Vercel

    Checks the build against the design and shares a preview link.

    Preview + notes
Feedback from users and the team becomes the next brief.

Tools I useClaude, Claude Code, Cursor, Figma, Figma MCP, Figma Make, Pencil, ChatGPT, Gemini, GitHub, Vercel

05 — Off hours

When I'm not designing

You may find me capturing dog photos on the street. And please, feel free to stop by to talk about dogs.

A curly black dog stretched out asleep on a patterned bench
A small tan dog in a green sweater in an office
A Shiba Inu in a red harness under a desk
A fluffy white dog curled up in someone's lap
A tricolor dog on a leash in autumn leaves
A black shepherd mix smiling on a patterned rug
A golden retriever in a sweater on a subway platform