MyCha

MyCha is a native iOS app that I took from a problem to a working app. I defined it, designed every screen, and built it in SwiftUI. Instead of focusing on reviews for others, it's more of a personal log that includes a journal, map, wishlist, pantry, and calendar.

I've had a matcha latte almost daily for the past 10 years, and always struggled to keep track of them. I was tracking the ones I bought in three separate notes, plus my camera roll and texts for the matchas I made at home. My problem was that rating apps list one score, grouped under "coffee & tea," to help you pick where to go, not really remember what you got. Drink trackers, on the other hand, either log only homemade coffee for caffeine content, or only café drinks to see on a map. I wanted to see my matcha across both, and remember what I thought of each one.

Since logging is the primary use case, the journal is the landing page, with a List / Map toggle for finding a drink by when or where I had it. The list, map, calendar, and each place's page all read from the same data, so nothing has to be logged twice.

Beyond logging, I wanted to remember where to go next, so the wishlist holds both places worth trying and favorites worth going back to. And since matcha powder loses its freshness once it's opened, the pantry tracks each powder, with freshness automatically calculated from the day it was opened.

The first build logged every drink I bought: milk tea, fruit tea, coffee. I scoped down to matcha and rethought the information architecture and visual design throughout. Drag to compare the logging form and journal.

I drew each illustration by hand to give MyCha some personality.

Every screen draws from one set of tokens (colors, a type scale, spacing, and corner rounding) and reusable components.

Every text style scales with the reader's text size.

I built MyCha with Claude Code. For me as a designer, AI has allowed me to build in new ways, but it hasn't replaced craft. I came up with the design first, let AI take me there, and then continued to iterate with design principles in mind. AI is not the one deciding the design, it is simply a tool to empower it.

My first round of user testing revealed to me what I couldn't see as the only user. When testers attempted to log matchas, certain copy did not translate clearly. Furthermore, the map search seemed to do nothing because the map never moved. I rewrote the labels and made the map zoom to the results and list them as cards. I look forward to testing with more people to continue improving the MyCha experience.