Case Study · Consumer AI · Fashion Tech
Feed2Fit
A personal AI stylist that closes the gap between the outfits you save and the outfits you actually wear.
A quick read before you dive in
An AI styling assistant that turns creator-inspired looks into personalized outfit recommendations — matched to your own closet and personal style.
Personal style is a form of identity. The tools that exist today (Pinterest, Instagram) inspire but never convert. Feed2Fit bridges that gap with AI that actually knows you.
Fashion-conscious Gen Z and Millennial women (18–32) who follow creators for style inspiration but feel overwhelmed translating it into their own wardrobe.
As product designer & builder, I led research, prioritization, product decisions, and the end-to-end experience — from problem framing to the shipped flow.
The gap between inspiration and action
Gen Z spends an average of 2+ hours per day consuming fashion content — but converting that into actual outfits is overwhelming. Too many options, no personalization, no memory of what they already own. The discovery-to-wardrobe pipeline is completely broken.
People don't lack inspiration. They lack a bridge from inspiration to action. That's the problem Feed2Fit solves.
Save, don't wear
Inspiration piles up in saved folders and camera rolls — almost never revisited when it's time to get dressed.
No personalized styling
Feeds show what's trending, not what works for a specific body, palette, or wardrobe.
Closet blindness
Users forget what they already own, so they buy duplicates or ignore great pieces they have.
Content overload
Every platform pushes more — but nothing curates what's actually usable for one person.
Who are we truly building for?
Two distinct user groups, one shared friction — the gap between scrolling and wearing.
The Style-Curious Scroller
Consumes fashion content daily but struggles with the inspiration-to-outfit gap. Has a phone full of saved posts she never revisits.
- ·Curated looks without endless scrolling
- ·AI breakdown of why an outfit works (colors, silhouette, vibe)
- ·A way to see if she already owns something similar
- ·A personal space to save and organize looks she loves
The Content Creator
Creates style content and wants their looks to be shoppable and accessible to followers beyond the scroll.
- ·Follower engagement with their actual outfits
- ·Easy discovery format for their content
- ·Attribution and reach for their personal style
What the research kept telling me.
- ✦The friction isn't discovery — it's translation from inspiration to an outfit that fits the user's actual closet.
- ✦AI advice is only trusted when it references what the user already owns.
- ✦Creators want their content to live past the scroll, not vanish into a saved folder.
What hurts most — and why it matters
Mapped across urgency, frequency, and impact from user research.
| Pain point | Urgency | Frequency | Impact | Chosen |
|---|---|---|---|---|
| Can't recreate inspo looks | High | High | High | ✓ |
| No personalized styling advice | High | Med | High | ✓ |
| Don't know what they already own | Med | High | Med | — |
| Too much content, no curation | High | High | Med | — |
Close the inspiration-to-action gap. AI styling breakdowns combined with personal closet comparison is the killer feature combination — no one else does both together.
Four features. One seamless flow.
Discover Feed
Curated grid of creator-inspired outfit looks — browsable, beautiful, and organized by aesthetic and vibe.
AI Stylist
Conversational AI that breaks down any look — explaining colors, silhouette, styling principles, and how to make it your own.
My Closet
Personal wardrobe tracker where users catalogue what they own — enabling real-time comparison against new looks they discover.
Saved Looks
A personal moodboard and wishlist — the place where inspiration becomes an actual plan.
What I picked — and what it honestly cost.
| Solution | Benefit | Limitation |
|---|---|---|
| Discover Feed | High engagement, beautiful UX | Needs ongoing curation effort |
| AI Stylist | High value, deeply personalized | API-dependent, cost per query |
| My Closet | Unique differentiator, creates stickiness | Requires upfront effort from user |
| Saved Looks | Low friction, instant value | Passive — doesn't drive action alone |
Ship all four features as one connected loop: Discover → Save → Compare with Closet → Ask AI Stylist. Each screen makes the next one more useful — the value is the loop, not any single feature.
How I moved from problem to shipped product.
- Step 01Behavior auditStudied save-vs-wear gap in Gen Z fashion→
- Step 02User archetypesStyle-curious scroller + creator→
- Step 03IA designDiscover, Save, Closet, Stylist→
- Step 04Flow prototypeLoop between screens mapped end-to-end→
- Step 05AI integrationStylist grounded in user closet→
- Step 06Screen build5 core screens designed and built→
- Step 07ShipInteractive prototype deployed live→
- Step 08NextAdd outfit history + wear-count tracking
What was actually built and shipped.
"Feed2Fit proves that AI can make personal style genuinely personal — not just algorithmically optimized. It's fashion built around you, not around engagement metrics."
Snapshot metrics above describe product features and market context — not verified adoption numbers.
What this project taught me as a PM.
AI needs personal context to feel personal
A stylist that doesn't know your closet is a search engine. Grounding recommendations in owned items is what makes it feel like yours.
Design the loop, not the feature list
Individually, each screen is fine. Together, they change behavior — that's the product.
Fashion tech doesn't have to look like commerce
The scrapbook / lookbook feel earns time on screen in a way product-grid UIs don't.
What I'd test next
Measuring how often a saved look becomes a worn look — the one metric that proves the loop actually closes.