We implemented a design system across two code bases, a code sandbox, and Figma. The goal was to establish clear guardrails and standards for craft, while we scaled the design team from one to three.
At Series A, product-market fit came from shipping fast. With growing pains, systems means going even faster, but with more precision.
No more frankensteined UI, QA bottlenecks, or subjective design debates.
Plus a satirical tone to the content. I was challenged with implementing this into a trustworthy fintech app.
Pay the bills you already have and collect rewards on every one of them.
Rewards are issued by the card network, not by Chexy.
We now spend less time redesigning core components every new feature. A third the time on visual QA. And continue to improve our system with new enhancements. The best part is our non-technical team can now prototype an experiment without the coordination costs.
We now have more time to give care to those moments.
Choose a colour, a title and optional partner mark. This works for our same-day turnaround times. Now, personalized promotional cards drive an 18-26% clickthrough.
You’re in with 400 entries. Invite friends to add more.






Handoff.md sets the defaults in the repo, so every new sandbox starts with the right type, colour and components. Lints stop anyone breaking them. Weekly design reviews catch what’s missed and get the team fluent in it.
Running list of decisions to carry into the Linear tickets. Add to it with every related PR.

Months of groundwork came first: the product sandbox I had started long before this, which meant the three weeks weren’t starting from zero. The rebrand itself came with organizational buy-in.
I studied Airbnb, Uber and Wise. AI carried the implementation spread in code — without it, end to end across two codebases, the sandbox and Figma could not have happened with such velocity.