AstroSight
01 · The Problem
AI astrology has a trust problem before it has a design problem — people don't believe a prediction app unless it looks and feels credible first. That was the actual brief here, even though nobody phrased it that way.
02 · The Process
I built one design system and used it across three surfaces — mobile, web, and the marketing site — instead of treating them as separate projects.
Mobile got 25+ screens, web got 30+, the marketing site got 20+.
The hard part wasn't any single screen. It was keeping the same visual language holding up whether someone's looking at a daily reading on their phone or a full chart on desktop.
03 · Key Contributions
Designed 25+ mobile app screens covering daily predictions, chart readings, and account flows
Designed 30+ web app screens extending the same system to desktop
Designed 20+ marketing site screens to convert visitors into users
Established a single cross-platform component library used across all three surfaces
04 · The Outcome
Shipped all three, consistent system across the board. Not flashy work, but it's the kind of consistency that actually makes an AI product feel less random and more trustworthy.
Selected screens — placeholders