UONEX AI
UONEX AI.
A conversational shopping platform for Indian malls. Shoppers describe what they want in their own words, slang included, instead of walking a directory store by store.
- 60%Less time to discover a product
- 3xEngagement and session duration
- +40%Onboarding completion
My role
Founding designer, and the only one.
No design team, no handoff templates, no design system and no prior research. In four months from research to App Store launch, every decision also had to be buildable by a small team, fast.
- 01
Research and IA
What Gen Z shoppers actually do in a mall, and the structure that follows from it.
- Research and information architecture with no prior user research to build on
- Dual-world architecture: AI World for intent, Browse World for exploration
- 02
Product design
Every screen, from onboarding to store profiles, built without an existing system.
- Every product screen, with no existing design system
- 03
Brand and launch
The identity, the marketing system and the physical touchpoints in the mall.
- Full visual identity and marketing system for the Indian market launch
- In-mall QR signage and physical placement design
- Worked with
- The founder and a 3-person engineering team
- Scope
- Everything from the first user flow to the in-mall QR signage
- Status
- Live on the App Store and Play Store
Impact
Less time to discover a product
60%
We stopped making users speak algorithm.
Engagement and session duration
3x
Shoppers stayed when the mode fit.
Onboarding completion
+40%
A real result in the first 30 seconds.
Satisfied with recommendations
70%
From post-interview feedback.
Clients landed on the app design
20+
Won on the strength of the product.
Why do shoppers leave empty-handed?
Malls have everything. Finding it is the problem.
The problem
Malls have everything, but shoppers still leave empty-handed.
Hundreds of stores, and no way to find what you want quickly. Discovery meant memory, luck, or asking staff who may or may not know their own inventory.
- 01
No cross-store visibility
Nothing connected what a shopper wanted to which store had it, or where that store was.
- 02
Directories repeat the problem
Most apps answered with a directory, which is just a digital version of walking the mall.
- 03
Two minutes, then they leave
When shoppers can’t find something quickly, they don’t try harder. The outcome had to move from “gave up” to “bought it.”
Search, or browse?
Both, in one app, without a switch.
The core decision
Two worlds in one app: AI for intent, Browse for wandering.
Most discovery apps pick a lane, search-first or browse-first, and lose whoever isn’t in that mode today. I built a dual-world architecture so the app matches how people actually shop: the AI World answers a specific want, the Browse World supports open exploration, and shoppers move between them without noticing a switch.
shoppers never see a switch
How should it feel?
Like texting someone who knows every store.
Decision 01 · Discovery
One conversational exchange, in the shopper’s own words.
Shoppers used to touch three to seven stores before finding what they came for. Now they type what they want the way they’d say it out loud. The system reads intent, searches every store in the mall, and returns ranked results with exact locations.
Decision 02 · Modes
Moving between worlds without friction.
AI World handles intent, Browse World handles wandering, and both live in the same app. A shopper who came for one thing and started browsing never has to change apps or start over.
Decision 03 · Onboarding
Onboarding that demonstrates instead of explains.
Users opened the app, couldn’t tell what it did, and closed it. I rebuilt onboarding around one real query and an actual result, so by the time setup finished, users had already used the core product.
Decision 04 · Store profiles
Mapped to how people actually move through a mall.
Getting from "the AI found it" to "I’m standing in front of it" had to be frictionless. Store profiles show real-time availability alongside in-mall directions.
Browse World
The mall, in your pocket.
Before anyone typed a word to the AI, there was the Browse World: the way in for shoppers who came to wander. Malls, stores, brands and products, laid out in the order you’d walk them, with editorial type, warm tones and one question answered on every screen: where can I get this near me?
From the mall to the cardigan
01Home
Start with who you’re shopping for, and what’s new nearby.
02Mall
Every mall gets a profile, opening hours and a store directory.
03Store
A store’s hours, rating and directions, then its products.
04Brand
Brand pages with followers and what’s in store right now.
05Product
Colour, size and price, with offers in view.
06Nearby
The store it’s in, and the others close by that have it too.
07Editorial
Collections and brands to explore, set like a magazine.
Trade-offs · where it landed
Four months of picking a side, or refusing to.
Where most apps commit to one answer, UONEX kept both, and spoke the shopper’s language.
Small dots are what we considered. Big ones are where it landed.
Behind the work · 21 notes
A wall for a product that didn’t exist.
Four months, no research to build on and no design system, so everything started as a note.
- Malls have everything
- Shoppers leave in 2 minutes
- 3 to 7 stores per search
- A directory is the same problem
- Talk like a friend
- Slang is fine
- Gen Z, busy mall, no patience
- AI World for intent
- Browse World for wandering
- No mode switch
- Onboard with a real query
- First result in 30 sec
- Found it → standing in front of it
- Store map, not a database
- Real-time availability
- Ranked results + location
- No design system yet
- Small eng team, ship fast
- QR signage in the mall
- Brand for the India launch
- Social content for launch
What went beyond the app?
The brand was working before anyone opened it.
Beyond the app
The first impression was already working before anyone opened the app.
- 01
Visual identity
A visual identity designed for the Indian launch.
- 02
In-mall QR signage
Physical placement that turned foot traffic into organic downloads.
- 03
Launch and social
The marketing system and social content that carried the brand into launch.
Conclusion:
What this project was really about.
Conclusion