<!———
Work info
———>
My Role:
Product Design Intern: Redesigned the web platform's information architecture, interaction patterns, and visual design across three core screens.
<!———
Context
———>
A mobile app running in a browser window
Shuk is a two-sided rental platform that connects small landlords with tenants. Landlords manage one to a hundred units through Shuk: rent collection, maintenance requests, tenant screening, lease management. Tenants use it to find and apply for rentals, message landlords, and track their lease. The mobile app handled all of it well.
When I joined as an intern, the web platform was that same mobile app displayed on a desktop screen. Nothing changed between phone and browser. Filters lived behind a full-screen modal. The landlord dashboard ran through a sidebar menu with nine links, each going to its own page. The messaging screen crammed three panels into a width that barely fit two.
Every feature existed. None of it had been rearranged for the screen it was on.
Redesign Outcomes
150+
properties onboarded on the platform
2x
faster dashboard navigation, based on usage by the CEO and five landlords
Web users aren't doing what mobile users are doing
A tenant scrolling Shuk on their phone is browsing casually, maybe on the bus. A tenant on their laptop is comparing apartments, with tabs open to Zillow and Apartments.com next to Shuk. They need to filter fast and scan a lot of listings at once. The same gap applied to landlords on desktop, who sat down to manage their full portfolio, not just check in.
<!———
Audit
———>
What I found: buried filters, scattered navigation, cramped messaging
I audited the existing web platform against Zillow, Apartments.com, and Airbnb. The web version mirrored the mobile app screen for screen, with navigation and density built for a phone and interaction patterns that added steps instead of using the available space.
<!———
Testing
———>
The CEO is also a landlord
I tested designs with the CEO and CTO directly as I didn't have access to end users during the internship.
The CEO's involvement made up for a narrower pool in one specific way: he wasn't evaluating mockups as a stakeholder approving a designer's work. He manages properties every day and would use this dashboard the next morning. When he said the old dashboard was too nested, that wasn't an aesthetic preference. He felt it every time he checked which tenants had paid.
His feedback drove the biggest structural decisions in the redesign.
<!———
Design
———>
Redesigning for web
The listing page
The competitive analysis showed that Zillow, Apartments.com, and Airbnb all keep the broadest filters visible inline: location, dates, price, size. Tenants start broad and narrow down, so those belong at the top level. Everything else is a refinement that can live behind a secondary panel.
I moved the grid from two columns to four. On a desktop, being able to scan more listings at once changes how people compare.
The landlord dashboard
I restructured the dashboard around what a landlord checks first. Funds owed, recent payments, lease end dates, and their rating sit at the top of the page before any scrolling. The nine sidebar pages collapsed into a single overview with sectioned content, so a landlord can check the state of a property without clicking into a sub-page.
Messages
The old messaging screen had no lease context. If a tenant needed to check a lease date or rent breakdown mid-conversation, they had to leave and find it elsewhere. I added an "About your lease" panel accessible from the conversation header, so the context a tenant needs sits alongside the thread.
The redesigns entered the development pipeline during my internship and shipped after my internship.
<!———
Learnings
———>
What I learned
Involved stakeholders change the work
Oliver and Miles were approachable and willing to critique the designs honestly. They were transparent about how they used the platform themselves and shared the data coming into it, which helped me prioritize what information to surface on each screen. Having stakeholders that involved made a real difference in a project where I couldn't reach end users directly.
