Finding a roommate is left almost entirely to chance.
RoomieMatch began with my own problem. Looking for a place near campus, I wanted to live with people who shared my habits and interests — and realized there was no real way to find them. Students end up piecing it together through WhatsApp groups, Facebook channels, and one-off DMs, then hope the person they move in with turns out to be compatible.
Three problems sat underneath that:
Imagine if students could choose their roommates and ideal location, instead of leaving it all to chance.
Who I designed for — and the questions that drove it.
I surveyed 15 students who had lived with roommates before, alongside informal interviews. Three findings shaped everything that followed: lifestyle mismatches are the leading cause of roommate conflict; people want transparency but fear awkward conversations early on; and most students rely on gut instinct rather than any structured way to compare.
Roommate conflict is rarely about personality. It's about unspoken expectations.
I built the project around a primary persona drawn from the people living this problem: budget-conscious university students who want somewhere affordable near campus and a roommate who respects how they live.
"As a college student, I want to find a roommate who respects my study schedule so that I can focus on academics without distractions."
— RoomieMatch persona — Sarah, 21, Student
Sarah's goal was simple — a responsible, study-friendly roommate near campus. Her pain points were just as clear: past roommates who were noisy and unreliable, and the difficulty of finding an affordable place close to the university. She's comfortable with mobile apps and social media, so a phone-first product fits how she already lives.
That framing turned into three "how might we" questions the whole design had to answer:
The bets I made — and what I left out.
As a solo concept, I kept the scope to the core loop — set up who you are, find compatible people, talk safely — and made three deliberate bets:
Compatibility before browsing. Capture lifestyle, budget, and location up front so matches mean something, rather than dropping people into a cold list of rooms.
Make discovery feel good. Borrow the swipe interaction people already know from dating apps, so finding a roommate feels fast and a little fun instead of like filing paperwork.
Build trust into the product. Verification, profiles, and in-app messaging replace the unvetted WhatsApp-and-DM scramble students fall back on today.
The marketplace machinery — payments, lease paperwork, dispute handling. Those matter for a real launch, but they'd have buried the one thing I wanted to prove: that compatibility-first matching is worth using.
One guided loop: set up, match, and talk.
RoomieMatch turns roommate-hunting into a single, calm flow. You tell it how you live, it surfaces compatible people and places, and you get to know them inside the app before anything is decided.
Chance pairings fail because nobody captures fit before move-in.
A short, guided setup asks what you're looking for, your budget and timing, and the lifestyle signals that actually predict friction — early bird or night owl, neat or casual — then feeds them straight into matching.
Three design decisions carried the experience:
Designed around lifestyle, not identity
Instead of leading with photos and bios, the experience centers on habits and living preferences — shifting the decision from "Do I like this person?" to "Can I live with this person?"
Made compatibility visible and comparable
A compatibility score and side-by-side comparison highlight alignment and friction points, so users decide faster and more confidently without reading long profiles.
Reduced emotional friction during onboarding
The flow frames questions as neutral and practical, encouraging honest answers without feeling judged — which raised completion and trust in the results.
Where it stands — and what I'd test next.
RoomieMatch is a concept I took from a personal frustration all the way to a hi-fi, end-to-end prototype: onboarding, matching, comparison, community, and chat. It's the project where I learned to start from a real problem and a specific person, then let that drive every screen.
It hasn't been formally usability-tested yet, but I shared the prototype with a few students to gut-check whether it actually solved the problem. One reaction stuck with me:
"Why didn't I have this kind of app when I came to Rochester? It would have solved my problem of finding a place and good people early on."
— Aryan, student who relocated to Rochester
That's the signal I was looking for at this stage: the problem is real and the concept resonates with the exact people it's for. The next step is moving from "this resonates" to "this measurably works."
If I carried it forward, I'd put the swipe-matching model in front of real students first, since it's the riskiest bet: does compatibility-based swiping actually beat browsing listings? From there I'd validate which lifestyle signals genuinely predict a good match, then design the verification flow for real — trust is the feature that makes everything else safe to act on.
Why this project matters
It reflects how I approach UX: start with human discomfort, reduce friction through structure, and design systems that help people make better decisions — not just prettier screens.