01 / 04
Slab IQ
A ground-up UX and visual redesign of Slab IQ, a trading platform for sports card collectors — replacing a rushed go-to-market MVP with a real design system, legible data, and working buy/sell/trade flows.
Consistent design system and clear data visualization where there was none — pricing and card data are now legible enough to actually inform a trade decision.
Buy/sell/trade flows and browsing rebuilt from broken and confusing to functional, ahead of the product's public go-to-market.
Slab IQ
Problem
- Slab IQ's initial build shipped fast without a real design system — inconsistent patterns and no clear visual hierarchy across dense, cluttered layouts buried the pricing and card data collectors actually needed to make a decision.
- The core buy/sell/trade flows were broken and hard to complete, and browsing the card catalog itself was confusing enough to lose new users before they ever got to a trade.
Approach
- Built Slab IQ's design system from 0 → 1 — tokens and components that gave the product a consistent visual language and enforced hierarchy instead of ad-hoc styling per screen.
- Redesigned the core buy/sell/trade flows end to end, and reworked browsing and catalog navigation to make finding and evaluating a card straightforward.
- Rebuilt the product's data visualization so pricing history, card value, and comps are actually legible at a glance.
- Set visual and brand direction, and weighed in on product strategy and positioning as Strategy Consultant, not just the interface.
Solution
- Shipped a redesigned platform: a real design system, decluttered and hierarchical layouts, legible data visualization, and functional buy/sell/trade flows in place of the original MVP.
Impact
- Consistent design system and clear data visualization where there was none — pricing and card data are now legible enough to actually inform a trade decision.
- Buy/sell/trade flows and browsing rebuilt from broken and confusing to functional, ahead of the product's public go-to-market.
Findings
- Product currently in Beta testing, pre-G2M
02 / 04
Pacific v2.0
Design strategy for Pacific's next-generation library — a slots-based component architecture and cross-platform prop-naming system built to scale with SoFi's product surface, not against it.
Fewer component variants to maintain, with more flexibility per component.
Faster, better-scoped migration tickets thanks to the Cursor-generated hand-off docs.
A shared naming vocabulary across design, web, and native that removes a recurring source of cross-platform miscommunication.
Pacific v2.0
Problem
- Pacific v1's component variants had grown faster than the products they served — each new use case meant a new variant instead of a flexible base, and prop names diverged across design, web, and native, slowing every cross-platform build.
- Hand-off between design and engineering leaned on manual, ad-hoc documentation, so v1-to-v2 migrations were slow to scope and easy to under-spec.
Approach
- Designed a slots-based composition model for the native library, replacing rigid variant sets with flexible, composable component APIs.
- Authored a cross-platform prop-naming standard, aligning how design, web, and native each describe the same component properties.
- Built an AI-powered Cursor skill that generates a one-command v1-to-v2 comparison — proposed changes, requirements, and hand-off-ready ticket copy — cutting manual scoping work out of every migration.
Solution
- Refactored core components against the new slots model and naming standard, shipping updates directly into the Flutter codebase alongside engineering.
- Rolled the standards into a recurring grooming ritual with Mobile Core and native Pacific designers, catching scope and slot questions before they became rework.
Impact
- Fewer component variants to maintain, with more flexibility per component.
- Faster, better-scoped migration tickets thanks to the Cursor-generated hand-off docs.
- A shared naming vocabulary across design, web, and native that removes a recurring source of cross-platform miscommunication.
Findings
- Native v2 library rollout and adoption content in progress, ahead of a design leadership review.
03 / 04
Pacific Migration
Sole design lead for SoFi's legacy-to-Pacific migration — flipping the QA model from reactive to proactive across a large, parallel engineering effort.
Fewer VQA cycles and merge-request revisions per migrated screen.
Feature-team designers kept their capacity for roadmap work instead of absorbing migration QA.
Pacific stayed unblocked through the highest-demand period in the system's history.
Pacific Migration
Problem
- Migrating legacy screens to the Pacific Design System had been a reactive process — engineers implemented first, and design caught misused components and pattern drift during QA, after the work was already built.
- That cycle meant slow reviews and repeated rework, just as request volume from the migration and major launches was climbing sharply.
Approach
- Took on migration design solo, partnering directly with a large parallel engineering effort across more than a dozen work streams.
- Flipped the model: rebuilt each migrated screen in Figma with the correct Pacific components, patterns, and tokens before engineering touched the code, so implementation started from an already-correct spec.
- Triaged and absorbed the majority of incoming Pacific requests personally during the resulting demand spike, so migration work and feature launches didn't stall waiting on the design system.
Solution
- Delivered proactive, pre-built specs for every migrated screen, replacing after-the-fact QA with up-front correctness.
- Kept Pacific requests moving fast enough that concurrent priority launches shipped on schedule alongside the migration.
Impact
- Fewer VQA cycles and merge-request revisions per migrated screen.
- Feature-team designers kept their capacity for roadmap work instead of absorbing migration QA.
- Pacific stayed unblocked through the highest-demand period in the system's history.
Findings
- The proactive model is now being folded into Pacific's standard process for future migrations.
04 / 04
Credit Card L1 Redesign
A user-centered redesign of SoFi's credit card experience, translating customer needs into a more intuitive, visually cohesive flow.
A more intuitive, visually consistent credit card experience for members.
A model for user-centered redesign work reused on later product surfaces.
Credit Card L1 Redesign
Problem
- The existing credit card experience didn't clearly reflect how members actually thought about and used their card, creating friction in everyday moments.
Approach
- Partnered closely with product and engineering to ground the redesign in real customer needs rather than surface-level visual polish.
- Iterated on the core flow and visual language to make the experience feel more intuitive at every step.
Solution
- Shipped a redesigned end-to-end experience, delivered in close partnership with engineering to keep design intent intact through implementation.
Impact
- A more intuitive, visually consistent credit card experience for members.
- A model for user-centered redesign work reused on later product surfaces.