independent, self-initiated
product design case study
DAILY
DRESS
ME
Weather Trust System
2026 — Product Design
Independent redesign proposal based on real user feedback
I gathered while working with the DDM team.
Role: Product Design (UX + Interaction + System Design)
Tools: Figma, Claude AI, Weather API logic mapping
Deliverables: UX logic system, wireframes, prototype,
visual direction
01 — The Problem
"The app
recommended
ballet flats.
It was raining."
While managing DDM's social channels, I had a direct line
to user sentiment. The same complaint kept surfacing: outfit recommendations that ignored real conditions. This wasn't
a hypothesis — it was a pattern I watched form in the comments.
02 — The System
TRUST FRAMEWORK
user
location
weather
api
filter
engine
outfit
database
4 — SEASONS WERE THE ONLY INPUT. ZERO REAL-TIME CONDITIONS.
5 — Filtering rules implemented to restore user trust
The fix wasn't visual — it was architectural.
A filter engine sits between the weather API
and the outfit feed, applying conditional rules
before anything reaches the user.
filtered
feed
01
02
03
04
05
RAIN ≥ 60%
Remove open shoes— sandals,
flats, mules
Temp < 10°C
Boost high layering outfits
to top of feed
Promote rain-tagged items
to top of feed
Rain ≥ 40%
Exclude flowy
and lightweight fabrics
Wind > 25 km/h
Exclude coats
and heavy wool entirely
Temp > 25°C
03 — The Solution
WIREFRAMES
Three screens that make the logic visible — not hidden infrastructure, but a trust layer users can actually read.
01
Confidence Tags
Weather-matched labels replace generic outfit names. Users see exactly why an outfit was recommended.
02
Swap Mechanic
Rather than hiding or removing outfits, the system adapts them. Users see what changed and why — building trust through transparency.
03
User Controls
Users can override the logic manually. A summary card shows exactly which rules are active —
no black box.
04 — In Action
prototype
The swap mechanic is best understood in motion.
The video shows the full user journey from weather detection through outfit adaptation — including the strikethrough swap panel and confidence indicator
in real time.
Watch: swap mechanic in action →
05 — What's Next
next steps
01
User
Testing
Validate that the swap panel builds trust rather than creating confusion. Measure return rate before and after.
02
swap
library
Build the full modular outfit database — every outfit tagged with swap alternatives for rain, cold, and wind conditions.
03
trust metrics
Define KPIs around trust — session length, repeat opens, and user-reported confidence score after outfit selection.
06— Case study closing
WHAT THIS PROJECT REMINDED ME
A product that loses user trust is a product that loses everything. This redesign is a reminder that users must always sit at the centre of every decision — not as an afterthought, but as the starting point. Their needs shift, their context changes, and products that fail to adapt
get abandoned. Trust isn’t a feature you ship once.
It’s something you continuously earn.
“Trust isn’t a feature
you ship once.
It’s something
you continuously earn.”