Katy is my ideal design partner. She balances strong vision with pragmatism and takes the time to understand how the system works end to end. She is curious, detail-oriented, and genuinely invested in improving both the product and those around her.

Abe Kim

SOFTWARE ENGINEER, DOORMAT

Katy is my ideal design partner. She balances strong vision with pragmatism and takes the time to understand how the system works end to end. She is curious, detail-oriented, and genuinely invested in improving both the product and those around her.

Abe Kim

SOFTWARE ENGINEER, DOORMAT

Mailchimp

Product Designer (Agency) • 2020

I worked on an internal initiative supporting Mailchimp’s Wink design system as part of an agency team assembled by Superfriendly. We partnered with brand and product stakeholders to explore how Wink could scale a shared system beyond components.

MY role

Constraints

Wink was recently launched, so this initiative focused on extension and adoption. The platform needed to support cross-functional alignment and help Peeps (the name for Mailchimp teammates) feel more inspired and connected creatively.

Explorations

Reframing “Mailchimp-y” as a Playground

Mailchimp’s internal documentation leaned editorial. We explored a broader spectrum of expression and pushed toward a playground model instead of a static reference site. The goal was to create a digital space that invited participation and experimentation rather than prescribing how Mailchimp designs.During remote work in COVID, teams had lost in-office creative exchange. The platform aimed to reintroduce shared exploration in a structured environment.Once stakeholders saw that prototypes, archives, rejected directions, and system fundamentals could live in one place, they began referencing the Playground model in planning discussions.

Exploring IA and design references for the playground framework

Making exploration reusable with the Parts Bin

One concept we explored was the Parts Bin, a creative archive for ideas that did not ship but still carried value.
The intent was to treat exploration as cumulative rather than disposable, enabling teams to trace creative lineage, rediscover alternate directions, and reuse thinking across surfaces and campaigns.

Outcome

The engagement concluded before full implementation, but it clarified how Wink could evolve beyond components into shared infrastructure.It established a model for thinking about design systems as a living system rather than static documentation.

Reflection

Design systems scale with interpretation and support from internal culture. The most valuable work was not defining new components, but designing a space where teams could explore, learn, and align within shared constraints.

Taste the City

Advisor, Product Design • 2023-2026

Taste the City is a hospitality marketplace connecting restaurants, curators, and diners through SMS-guided food tours. Content creators curate routes to monetize their recommendations, users book surprise tasting tours, and restaurants get free exposure to new customers.When I started working with the team, their product was live in 100+ restaurants across Canada, generating strong traffic from their social content, but sales lagged. The team was about to fundraise and needed to improve conversion without rebuilding the MVP.

Challenge

The UX wasn’t smooth, so users dropped off while browsing routes and during booking. Visual hierarchy was inconsistent, imagery was deprioritized, and routes lacked enough context to build trust. The on the website UX didn’t match the premium pricing.

Mapped the end-to-end lifecycle to find the highest leverage opportunities to improve conversion and retention

FOCUS AREAS

I focused on clarifying value, tightening the booking journey, and improving conversion within the existing MVP’s technical constraints.

1. Route-First Browsing
Shifted from city-first navigation to route-first browsing to reduce cognitive load and make curated experiences easier to compare.2. Trust & Expectation Setting
Improved key touch-points and microcopy without introducing major engineering changes, balancing growth goals with MVP limitations.3. Partner Enablement
Created physical and digital touchpoints to reduce confusion, support onboarding, and protect the guest experience at scale.

Created a simple printed guide so restaurant partners could confidently deliver the TTC experience

Outcomes

Seam Social

Co-founder & Design Lead • 2021-2023

Seam Social is a customizable social platform and SDK that lets users extend identity across games and online communities.I designed the core product and mini app SDK, enabling users to create personalized mini apps, apply themes, and embed their profiles outside the platform.

How might we hack the cold start problem and get distribution for a new type of social media profile?

All users set a profile song during onboarding

One of many community-requested profile themes

CHALLENGE

Seam faced two cold-start problems: a standalone social platform without users, and an SDK without distribution to attract developers.Without users, developers had no reason to build. Without compelling features from developers, users had no reason to stay.

APPROACH

To address the cold start on both the platform and the SDK, I focused on distribution. We built a lightweight integration that embedded Seam profiles and mini-apps directly into Pixels, seeding real usage and attracting developers through a single high-signal launch.

OPPORTUNITY

Pixels is a top blockchain MMORPG with over 250k DAU ( at the time) and an active social layer, but lacked flexible in-game identity features. This made it an ideal environment to introduce Seam, where profiles could be immediately visible, useful, and socially legible.

Outcomes

Embedding Seam identity directly into Pixels was our strongest go-to-market outcome. The integration also served as a distribution channel for the SDK, leading 12 external developers to build custom features.

Seam had an immediate 10x increase in DAU

Reflection

This project reinforced that distribution-first design beats building standalone for niche products. Embedding Seam's identity into existing social environments through lightweight SDK integration validated demand and focused engineering effort on what actually mattered.

Doormat Wallet

Founding Product Designer • 2024-2025

Doormat was a self-custody multi party compute (MPC) crypto wallet, leveraging threshold signature signing and distributed key generation. It later became a trading terminal with transaction batching and automation tools for advanced users.I joined as the first product designer as the team prepared to launch the initial MVP. This case study focuses on the key product and design decisions that shaped the path towards product market fit rather than any single product design process or challenge.

MAY 2024

I designed the MVP for our self-custody MPC wallet available on desktop, mobile, Chrome and iOS

Objective
Let users access any key from any of their devices using just a master password and another device to grant permission. This is a unique offering because it allows users to access their assets without entering a seed phrase (or recovery phrase) in new devices.Solution
We built a desktop-first, mobile-friendly web app alongside an iOS app and Chrome extension. Using custom-built MPC key management (not Turnkey or Privy) users could access any EVM or Solana key from any device without seed phrases or custody tradeoffs.Impact
This validated Doormat’s technical foundation and differentiated it from traditional wallets, but also revealed that security primitives alone were not enough to drive repeat usage.

OCTOBER 2024

We introduced key orchestration - the Keychains feature - to give power users an execution advantage

Objective
Create a differentiated trading advantage inside the wallet for advanced users.Solution
I led the design of the Keychains feature, a system that lets users generate and manage up to 100 wallets as a single execution unit and submit coordinated transactions across them in parallel.This enabled traders to distribute risk, automate strategies, and execute complex positions without external tooling.Challenges
Keychains introduced significant reliability risk: individual wallets frequently failed mid-transaction, breaking batch execution.I worked closely with engineering to map failure modes, model recovery paths, and design transparent error handling. We shipped a resilient fallback system with granular failure diagnostics and a “Retry Transaction” flow, allowing users to recover partial executions while infrastructure reliability improved.This approach reduced abandoned trades and preserved user trust during periods of network instability.Impact
Keychains repositioned Doormat from a secure wallet to trading execution infrastructure. It became the primary onboarding path for advanced traders and a core driver of power-user retention.

The redesign that Katy led, elevated our product in a way that immediately resonated with our community. The brand and visual direction finally felt aligned with who we are. Her ambition and sense of urgency pushed the entire team to move faster and aim higher.

Trevor Aron

FOUNDER, DOORMAT

February 2025

I led a full product redesign from an MPC wallet into a trading platform modeled after the Bloomberg Terminal

Objective
Support faster and more efficient trading workflows without users needing to rely on external tools or navigate outside of the product.Solution
I designed a unified, Bloomberg-style terminal that directly integrated Pump.fun, news, price tickers, live token charts, and a command line interface helping users minimize data fragmentation and execute faster.Impact
The Pump.fun integration became the primary entry point for over 60% of users, lifting 30-day retention to ~50%.

MAY 2025

We enabled automations for trade execution during peak market volatility so users could get some sleep

Terminal-native automation replacing manual scripts, reducing technical barriers without sacrificing control

Objective
Allow users to trade effectively even when offline during fast-moving markets.Solution
We built terminal-native automation rules, ranging from basic protections like stop losses to trading-specific logic such as round-trip protection, replacing brittle custom scripts without hiding complexity.Impact
Automations improved retention among power users and reduced reliance on external tooling, contributing to a ~50% lift in 30-day retention post-launch.

Katy is my ideal design partner. She balances strong vision with pragmatism and takes the time to understand how the system works end to end. She is curious, detail-oriented, and genuinely invested in improving both the product and those around her.

Abe Kim

SOFTWARE ENGINEER, DOORMAT

Outcome

Doormat evolved from an MPC wallet into a full trading terminal supporting Solana and EVM chains, Pump.fun integration, automated trading, and Keychains. The product supported 616 alpha users, 8,000+ wallets, $1.9M peak AUM, and $900K in trading volume, validating that power users will adopt complex tools when they meaningfully improve speed, control, and execution quality.

Welcome to my bookshelf!

This is a selection of books I've enjoyed over the past few years. I've been trying to read more fiction (sci-fi especially). Please share your recommendations below!