Maya Chen — her stories, what the Mirror knows, and every frame shot for her.
Maya Chen
34 · Seattle
Fictional. Every photograph is generated.
User stories
Six journeys, three months
Day 0 to month 3. The system earns the right to say more at each one.
Maturity
Four rungs, one shopper
The interface shows less as the fact count grows. That contrast is the claim.
- Just getting started0 facts
Your profile gets smarter as you shop. You can also tell me what matters to you anytime.
- Learning4 facts
I know a little. Enough to stop asking you the same things twice.
- Established6 facts
I know your size, your heel tolerance and roughly what you like. Fit is still where I am least sure.
- Strong11 facts
I know enough to personalise most shopping missions. You can change anything I have learned.
At six months
Every fact says where it came from
This is what it answers when you ask why. All of it is reversible.
Size & fit
- US Women's 8, regular widthconfirmed
You told me directly, and it has held across four purchases.
- Avoid narrow toe boxeslearned · 2×
You returned two pairs — a metallic block heel and a pointed flat — both because the toe box was too narrow.
Comfort
- Low heel preferredconfirmed
You told me during Sarah's Wedding and have not contradicted it since.
- Comfort matters past about three hourslearned · 3×
On long events you have consistently traded style for comfort at the final decision.
Style
- Minimalconfirmed
You confirmed this after I saw it in three missions.
- Clean silhouettes, subtle metallicslearned · 4×
Your direction choices across four missions.
- Architectural shapesexploring · 2×
Sarah's Wedding and Work Dinner — an early pattern, not yet a preference.
Avoids
- Rhinestones, large logosconfirmed
You told me directly.
- Traditional stilettoslearned · 5×
Passed over in five missions without exception.
Budget
- $100–$175 for event shoeslearned · 4×
Four missions, all settling in that band.
Colour
- Black, cream, muted green, goldlearned · 6×
What you actually keep, rather than what you look at.
Outcomes
Kept or returned, not clicked
Why it can say “you returned two pairs for exactly that.”
- Metallic block heel, size 8returned
Sarah's Wedding — Too narrow
- Pointed leather flat, size 8returned
Work Dinner — Too narrow
- Cream ballet flat, size 8kept
Work Dinner — Worn 11 times
- Sage linen shirtdresskept
Spring refresh
- Cream wide-leg trouserkept
Work rotation — Worn weekly
- Ivory silk shirtkept
Work rotation
- Ribbed knit top, ecrureturned
Spring refresh — Fabric too sheer
- Black flat sandalkept
Summer party
- Champagne bias slip skirtkept
Gallery opening — Worn 6 times
- Cream leather mulekept
Work rotation
- Strappy heeled sandal, goldreturned
Summer party — Heel higher than listed
- Tan leather clogkept
Everyday — Worn most weeks
- Black packable rain shellkept
Rainy commute — Lives in her bag
- Sage ribbed cardigankept
Spring refresh
- Cream ballet flat, size 8kept
Everyday — Replacing the pair she wore out
- Linen shirt, whitekept
Portland trip
- Charcoal wool overcoatkept
Winter layers — Worn all winter
- Black relaxed blazerkept
Conference — Worn 14 times
- Silk blouse, oxbloodreturned
Conference — Colour read brown in daylight
- Charcoal knit midi dresskept
Holiday party — Worn 9 times
- Black evening dresskept
Holiday party
- Sequinned mini dressreturned
Holiday party — Not her — too much surface
- Black lace-up derbykept
Work rotation
- Brown suede Chelsea bootkept
Autumn commute — Worn most weeks
- Olive quilted rain parkakept
Rainy commute — Worn 25+ times
- Merino beanie, blackkept
Winter layers
- Wool blend scarf, charcoalkept
Winter layers
- Cropped puffer, silverreturned
Winter layers — Too shiny in daylight
- Grey trail runnerkept
Weekend hikes — Worn every weekend
- Fleece midlayer, sagekept
Cabin weekend
- Wide-leg cord trouser, rustreturned
Cabin weekend — Ran a full size large
- Black fine-knit turtleneckkept
Winter layers — Worn weekly
- Oatmeal merino crewneckkept
Winter layers — Worn 18 times
- Black ankle bootkept
Autumn commute — Worn most weeks
- Stone belted trenchkept
Rainy commute — Worn 20+ times
- Black rubber rain bootkept
Rainy commute — Worn most weeks
- High-rise straight jeankept
Everyday — Worn constantly
- Cropped straight jean, ecrureturned
Everyday — Waistband gapped
- White low-profile sneakerkept
Everyday — Worn most weeks
- Black leather loaferkept
Work rotation — Worn 30+ times
- Navy work trouserkept
Work rotation — Worn weekly
- Structured shift dress, navyreturned
Work rotation — Shoulders too boxy
- Black slingbackkept
Birthday dinner
- Sage satin midi dresskept
Birthday dinner — Worn 7 times
- Metallic gold sandalkept
Summer party — Worn 5 times
- Low block sandalkept
Summer party
- Cross-body bag, tankept
Everyday — Daily carry
- Cotton tote, naturalkept
Everyday
Closet
30 items the market has to beat
Searched before the market, every mission. The $0 answer can win.
Black leather loafer
White low-profile sneaker
Metallic gold sandal
Black ankle boot
Cream ballet flat
Black slingback
Sage satin midi dress
Navy work trouser
Black evening dress
Brown suede Chelsea boot
Black rubber rain boot
Grey trail runner
Cream leather mule
Black lace-up derby
Tan leather clog
Black flat sandal
Stone belted trench
Olive quilted rain parka
Charcoal wool overcoat
Black packable rain shell
Oatmeal merino crewneck
Sage ribbed cardigan
Black fine-knit turtleneck
Charcoal knit midi dress
Black relaxed blazer
Ivory silk shirt
High-rise straight jean
Cream wide-leg trouser
Champagne bias slip skirt
Sage linen shirtdress
Occasions
Three, pulling against each other
One event can’t prove the shortlist reasons. These invert: outdoor against indoor, dressy against all-day walking.
- Dinner with Anna1 need
Friday · Seattle · Casual · indoor — Shoes
- Sarah's Wedding1 need
September 19 · Hudson Valley, NY · Semi-formal · outdoor — Shoes
- Chicago Trip3 needs
October 3–6 · Chicago, IL · Smart casual · indoor — Walking shoes, Dinner shoes, Coat
Imagery
Every frame, generated
Higgsfield — soul_2 for product, nano-banana for anything with Maya in it. No scraped photography, no real brands. A product with no shot falls back to drawn SVG.
Four frames, verified against each other by eye rather than assumed.
Profile
Chicago — coat and flats
Her closet
Friday dinner
The sage midi with each candidate shoe — the decision surface.
The dress she owns
Safe
You
Push it
Use what you own — $0
Gold sandal
Statement
Gold sandal, alt
Where the wedding actually happens — grass, not a studio.
The dress, as product
Hudson Valley
Ready
Ready, vertical
Ready, alt
Shot for a specific claim. The toe box exists because a review mentions it.
Toe box, from above
On the foot
On grass
Construction
“Less shiny” has to be visible, or the word did nothing.
Cooler
Polished
Brushed
Matte
What the market has to beat.
Sneaker
Ankle boot
Ballet flat
Navy trouser
Evening dress
Designed, not built
Two more shoppers
Specified in specs/ and not in the app. Maya is the only profile the code holds — these two exist as designs, and saying so here is cheaper than a page that implies three personas ship.
Nora Ellis
41 · Oakland · runs a small architecture practice
The same wedding, answered differently. Same catalog, same maturity, same hard filters — only the Mirror changes, so nothing else can explain the different shoe.
- US 9 wide, against Maya’s 8 regular — changes what qualifies
- High shine and ornament where Maya is minimal and matte
- Agrees on heel and budget, so the hard filters stay identical
- Owns nothing that qualifies — the closet pass returns nothing and says so
Dale Whitaker
52 · Seattle · buys for weather and for walking
Fit resolved without a history. He is a 10.5 in one brand and an 11 in another, so the fit pass has a real answer at a low maturity rung — which moves the fit story off the maturity ladder.
- Every mission is standing — no date, no dress code, no venue
- Two of the gate’s five dimensions have nothing to resolve
- Level 1, so he lands on “I’d start with these two”
- No portrait in the app by choice — the composed looks are Maya-only by construction