work

A design team of one, built to scale

I design across five Peek products, three actively. I automate the repeatable parts of my process so my time goes to product thinking, UX and taste.

Products: Leasing, Touring, Discover, Insights, AEO
Role: Sole product designer
When: ’25 → ’26
Tools: Figma, Claude Code, GitHub

5 productsdesigned as the only designer

3 AI skillsrunning ideation, handoff and QA

2 daysfrom mockup to production on AEO

Animated diagram of how I work. Slack, Gmail, Calendar, Jira, GitHub and my other tools flow into Claude, then Figma, then me, inside a loop of my design process: synthesis, PRD, usability testing, design system, handoff, QA and A/B tests.

The system

Systems handle the routine, so I can focus on the priority work that needs my attention.

→ 01

One hub

Jira, Slack, email, calendar, meeting notes and the roadmap, triaged into one morning view.

→ 02–04

Automated

Skills for ideation, handoff and QA, design system upkeep, first-pass research synthesis, and UI fixes shipped as pull requests.

→ 05

Where I step in

Customer interviews, product decisions, design reviews, cross-functional alignment and craft.

Stage01

One place for everything

Every tool in one view, so nothing slips between them.

Off-the-shelf tools couldn't handle work spread across five products, so I vibecoded my own work hub to prioritize my focus and make switching between projects seamless.

How it runs: Every morning an automated run syncs Jira, Slack, call transcripts and every other source, cross-references them with project status, and builds my agenda for the day.

In practice

Seven tools, one prioritized list. One run filtered out 31 items that used to land on my plate but needed nothing from me.

My work hub's Home view: today's agenda, items that need review, and what's coming up, pulled from Jira, Slack, Gmail, Calendar and meeting notes

Stage02

Skills that run my process

Standardized steps, so every project gets the same rigor.

Three AI skills cover ideation, handoff and QA, built on design methodologies like heuristics, accessibility and jobs to be done. AI also runs the first pass on research, synthesizing call transcripts and engagement data, so I start from patterns, not raw notes.

Checkpoints: Each skill pauses for my review before it moves on.

In practice

Every handoff and QA run follows one spec, and QA findings are drafted as Jira tickets, ready for my approval.

QA release console with my three skills, ux-project, handoff and qa-release, and a drafted Jira ticket queued for filing

Stage03

A design system built to scale

One source of truth, so the team can iterate fast.

Peek's products grew on brand guidelines set before I joined, layering legacy patterns with every launch. I audited them and rebuilt the design system around about 80 tokens and a full component library, documented so engineers and AI coding agents follow the same rules.

How I maintain it: Every change is versioned with its before state. New components are audited, then either added to the system or brought in line with it, and carried over to engineering in Storybook.

In practice

Every new screen starts from the same tokens and components.

Stage04

I ship the design work that waits

A designer who ships code, so UI debt doesn't wait for a cooldown that never comes.

In a lean startup, UI polish gets deprioritized and piles up as tech debt. With AI, I ship those changes as pull requests for engineering review. The biggest was AEO's building page: built for AI crawlers, it's now where prospective renters land as well, and needed a human-friendly redesign. With no dedicated designer or front-end engineer on that team, it sat in the backlog. I designed it and opened the PR so they could review it on real data.

Guardrails: My AI follows our engineering guidelines and code standards. Engineers review, merge and own the release.

In practice

Mockup to PR in one afternoon, in production two days later. One building's page went from 5.1 MB to 657 KB.

Before

After

Prospective renters now land here alongside AI crawlers, on a page that looks trustworthy enough to click through.

Stage05

Where I step in

Research, decisions and craft stay with me.

I run 1:1 customer interviews, make and own the design decisions, lead design reviews, partner with engineering, product and sales, and refine the details until they feel right.

How I decide: Every interview grounds my decisions in real customer workflows.

In practice

Portfolio View went from my first UX interview to launch in 7 weeks.

Desktop with a roadmap board, a Figma file, a work hub and a meeting window open side by side

What I learned

AI is only as good as the design and product knowledge behind it. It brings speed and consistency, but the product and UX calls stay mine.

Five products, one designer, and a process that keeps every release consistent.

Now that you've seen how I work, see it applied.

AI visibility reports · 6× ARR in ten monthsView project →

Portfolio overview of an AI visibility report on desktop, with a lifted recommendation card and a phone showing a score of 90.