← All work

Personal project

Wherart: all Parisian art in a single app.

A native iOS app that sorts Paris exhibitions by your taste and lets you visit them with others. I designed, built and shipped it alone.

Role
Founder, product designer and builder (solo)
Timeline
September 2025 to now
Status
Available on the App Store
Tools
Figma, Figma Make, SwiftUI, Supabase, PostHog, Claude

A feed sorted by what you like

Your art genres, favourite venues, location and time available decide what comes first. Distance and price sit on the card, so you can decide without opening it.

A map, one tap away

The feed is where you land. The map has its own tab, and clusters show how many exhibitions are in each area at a glance.

One screen with what you need to go

Distance, price, available visits, opening hours and address. I kept the information a visitor needs on the way there, and nothing else.

Visits with other people

Create or join a group visit: pick an exhibition, a date and a time, and set how many people can join. This is the feature that sets Wherart apart, and the one my beta testers struggled with most.

The problem

Finding an exhibition is easy. Deciding to go, and with whom, is not.

Exhibition information is spread across venue websites, listings and social media. Even when you find something, going alone is often the reason people don't. I wanted one place that sorts what's on in Paris by taste and makes the visit itself something you can share.

I also had a constraint of my own: I'm one person. Every feature had to earn its place against the time it takes to design, build and maintain.

My role and scope

Everything, with AI tools doing a lot of the typing.

I did the research, defined the product, designed the screens in Figma, built the native app and shipped it to the App Store. I used AI tools (Cursor, Claude, Figma Make) to write code faster. The product decisions on this page are mine.

Under the hood, in three sentences. I rebuilt the app from React Native to native SwiftUI. Exhibitions are synced every day from Paris Open Data into Supabase. An assistant powered by Claude answers questions about what to see.

Research

Early feedback told me people like the idea. It also told me where I was wrong.

I ran a feedback form on the first web version of Wherart, before the React Native and SwiftUI rebuilds.

4.5 / 5concept
4.4 / 5would recommend
4.3 / 5overall experience
4.4 / 5would use regularly

10 responses on the first web version and not the current iOS app. These scores say people are generous and interested. They don't prove demand.

4 to 5 of 10 testers had a problem with the group feature. It was not ergonomics. It was trust and understanding.

Their reasons: they hadn't tried it, they hadn't understood it, they didn't want to visit with strangers, or creating a visit failed. I had designed a feature for people who are comfortable meeting strangers around art, and most testers weren't.

Other signals worth acting on:

  • Two testers independently asked for a "notebook": exhibitions seen and to see, with a reminder before they close.
  • Art preferences didn't survive navigation, so people had to set them again.
  • One tester called the design "too sober" for an art app.

Decisions

Four calls I made, and why.

Onboarding cut to three screens

I looked at where people dropped off in PostHog and cut onboarding to three screens: welcome, art tastes, favourite places.

"Ending soon" badge at 25% of the time left

Few exhibitions qualify at any moment, but people who see one are the most likely to go. Low volume, high intent.

Accessibility and "kids friendly" filters, filled in by the community

Filters cover lift, adapted toilets, audio guide, guide dogs and kids-friendly, and users add this information on each exhibition.

An assistant in free text

You can ask "a photo exhibition this weekend near Le Marais" and get a short list of recommendations, in your own words instead of filters. It is live in version 1.6.4 and keeps your conversation history.

Welcome screen with a consent checkbox and a Continue button Onboarding: choose the art genres that move you Onboarding: choose the places you like to go out to
The three onboarding screens: welcome, art genres, favourite kinds of places. The app interface is in French.
Filters sheet with price, distance, duration, accessibility and waiting time Wherart AI chat greeting the user and asking which genre of exhibition they want Create a visit sheet with name, exhibition, date, time and maximum members
Filters with accessibility and waiting time, the Wherart AI assistant, and creating a group visit.

Results

What the numbers say, and what they don't.

200+installs (App Store and TestFlight)
15TestFlight beta testers
870+venues in the database
1,000+exhibition records processed by the daily sync

Installs are not retention, and I don't have retention data I'd trust yet. The beta tells me where people stall, and it shows me what to fix next.

What I'd do next

A pass, tickets in the app, Android, and a version for guides.

1. Wherart Pass (coming soon). Access to privileged events, private openings, skip-the-line tickets and guided visits.

2. Buying tickets in the app. Book directly from Wherart, without leaving the app.

3. Wherart on Android. So people can discover Paris exhibitions whatever phone they use.

4. A version for guides, to organise their visits.