Designing an app-based service from scratch to help university students form start-up teams.
Project overview
Team-up is a student-focused collaboration app designed to help university founders find the right partners to kick-off their entrepreneurial ideas.
By supporting structured matching, trust-building profiles, and community engagement, the app helps students gain confidence to start, build stronger teams, and reduce the number of ideas abandoned before they begin.
01. Understanding the users
During my time at NYU, many of our classes encouraged self-initiated UX group projects—an opportunity to take ideas beyond the classroom. I often found myself excited to turn these into something more real, something closer to a startup. But finding the right people to build with wasn’t easy.
Without peers to continue the momentum, those ideas slowly lost energy—and eventually faded. It turned out I wasn't the only one facing this issue.
I don’t know where to connect with other students who would like to create a startup.
One-on-one connections are good, but it’s hard finding dedicated people willing to put in the time and effort.
The tech stack requirement is always so high.
Everyone is looking for experience I don’t have yet.
After speaking with students and reviewing how universities currently support collaboration—from Slack channels for recruiting to social events—I found that few students felt these approaches helped them find the right collaborators. Three themes emerged:
As these patterns became clearer, three distinct user groups began to emerge.
Idea explorer
“I have ideas but they feel half-baked. I need someone to think through it with me first.”
Gain confidence before committing to a co-founder; validate ideas early before they go cold.
Ideas feel untested, solo work drains momentum, and many ideas never get started.
Team seeker
“I'm ready to contribute—I just don't know where I could find more start-up teams and their information.”
Join a startup team and contribute skills.
Active teams are hard to find, networks are limited, and reaching out can feel intimidating.
Team builder
“I have the idea and the momentum—I just need the right people to build it with me.”
Find reliable people across development, design, and marketing.
Interest alone is not enough; skill and seriousness are difficult to signal through cold outreach.
02. Defining the problem
Among the three user groups, I focused on the Idea Explorer—students interested in starting something but not yet actively building. A study found that 50% of university students fall into this category, yet only a small fraction go on to become team builders who actively build something.
Idea Explorers need more peer support to build confidence and motivation around early ideas. However, existing platforms (such as Slack) are dominated by people who are already building and recruiting team members. For students whose ideas are still taking shape, initiating conversations can feel hard and intimidating.
From my interviews and analysis with Idea Explorers, these are the key frustrations:
Based on these, I started generating ideas by asking:
How might we help students confidently share early ideas and form meaningful connections before they are ready to actively build?
03. Generating ideas & lo-fi
To address the key frustrations, I focused on creating a safe and encouraging environment for sharing early ideas, building trust through richer context, and fostering a sense of community through online interactions.
01
Use onboarding to help users identify their current stage and normalize being an Idea Explorer. Collect interests and preferences to surface more relevant connections.
02
Display student cards with key information for quick discovery. Create filters that help users find relevant connections.
03
Encourage detailed profiles that include interests, skills, goals, availability, and past experiences.
04
Design interactive event features that encourage participation before, during, and after events.
05
Topic-based spaces organize discussions around shared themes, so students can meet new members, share early ideas, and find peers around the questions they care about.
04. User testing
I recruited six potential users for a usability study and observed how they worked through four core tasks. The insights helped revise the design around the moments that needed more trust, clarity, or control.
01
During testing, one participant stopped partway through onboarding to ask why she needed to answer the questions. Others agreed a good profile mattered, but said there was too much to fill in at once, and they weren’t sure what a good profile should look like.
Onboarding became an AI-guided conversation. Instead of a static list of questions, the AI asks one thing at a time, explains why each answer matters, and responds to what users share, so it feels like talking to a person rather than filling in a form.
The AI explains why it asks, one question at a time
By the end of the conversation, the AI has enough to draft a complete profile, grouped into short, easy-to-scan sections. Users start from a strong example instead of a blank page, and only need to review, edit and publish.
Drafted from the chat, grouped into short sections
02
Event pages showed who was attending through demographic pie charts. Testers found the charts hard to read, and some said that if they didn’t fit into one of the categories, the event felt like it wasn’t meant for them.
I replaced the pie charts with simple bars that are easier to read at a glance. Attendees also became people rather than percentages: tapping an avatar in Who’s Going opens that person’s profile, so users can learn who will be there before deciding to go, and reach out to schedule a chat with anyone they’d like to meet.
Tap an avatar to view their profile and schedule a chat
Simple bars replace pie charts, easier to read at a glance
03
In the Hub, “Meet new members” was one tab among the discussion topics. Testers saw it differently: rather than a topic to browse, it was a way to welcome newcomers and encourage them to join in.
I moved new members into a dedicated section at the top of the Hub, shown as cards with each person’s stage, so newcomers are visible without competing with the discussion below. The section is collapsible, so users can hide it when they want to focus on the conversation.
New members get their own section, collapsible to focus on discussion
05. Refine UI
With the flows validated, I refined the visual layer: a palette with the right energy, and the components and screens students see most.
01
Lime and orange were finalized after exploring directions with AI. Together they create a warm, high-energy and youthful tone that matches the spirit of a student entrepreneurship app.
04
Refined the icons for greater visual consistency and accessibility.
02
Improved visual hierarchy and spacing so key information is easy to scan at a glance.
03
Both card types share the same core information. “Happening this week” cards are flyer-forward for browsing, while “Upcoming” cards show the time, location and RSVP so users can commit without tapping in.
06. Finished product
The finished experience brings onboarding, trustworthy discovery, profiles, events, and community together through one warm, high-energy visual language.
View live prototype07. Impact
To evaluate whether the design successfully creates a safe and welcoming environment for participation and connection, these are the metrics I would track in future usability testing.
| Metric | What it measures | Why it matters |
|---|---|---|
| Onboarding completion rate | Low-pressure environment | A high completion rate suggests the conversational AI format lowers the barrier to entry rather than feeling demanding. |
| Profile completion rate | Confidence to participate | A complete profile gives users an identity in the community and signals that the structured sections feel safe enough to fill in. |
| Event attendee info click rate | Informed decision making | A high rate means familiar faces or relevant peers help influence attendance decisions. |
| Message sent rate | Meaningful connections | Moving from discovery to conversation without friction shows the design is enabling real connections, not passive browsing. |
08. Reflection
Unlike my previous projects, I was the only researcher and designer for this one. That independence helped me stay focused and maintain consistency, but it also made narrow-mindedness and bias in design choices a real risk.
As a potential user of the product, I could inject too much of my own perspective and overlook other needs. For example, one tester felt that demographic attendee information could make people feel excluded if they did not fit the specified categories.
That feedback reinforced the importance of seeking diverse perspectives and staying open throughout the design process—especially when a product is meant to make people feel welcome.
