Perplexity is an emerging AI-powered search engine with 15M+ users, designed to deliver detailed, reliable answers. However, while it's actively growing, a lot of students still don't know what it is.
My team, Product Space at UCLA, worked with Perplexity to reimagine how Perplexity could improve student engagement.
THE RESULT
Starting From Scratch: How did I narrow down the problem space?
USER RESEARCH
Students + AI Tools: Diverse Uses, Different Goals
In order to gain product feedback on Perplexity and other competitors, we surveyed students at UCLA across 4 user types. With 135+ survey respondents and 30+ interviews across 35+ STEM & Humanities majors, we discovered that:
TEAM BRAINSTORM & EARLY IDEATION
Based on data, we identified themes and brainstormed a range of features, bringing them to discover which resonated most.
FEATURE OWNERSHIP
Designing Flow: Reducing Time Lost to Context-Switching Between AI Tools
After discussions with the design team, I took ownership of workflow efficiency and narrowed the problem space to a key question:
DESIGN EXPLORATION
Students were losing time switching between tabs — copying text into separate tools just to highlight or annotate.
To address this, I designed an inline annotation feature that allowed users to stay within the same workspace.
Plans Changed, Leading to Deeper Research
UNEXPECTED CHANGES
Additional user interviews changed my original idea. Students didn’t just want to annotate — they wanted a quick way to save quotes and insights directly in Perplexity.
FINAL DIRECTION
Instead of creating a separate tool, I built on Spaces, Perplexity’s collaborative folders, to become a lightweight text-saving hub.
USER STORIES & COMPETITIVE ANALYSIS
Discovering User's Frustrations with Tab Switching
90% of students spend 1-7 hours weekly on research, but always have to switch contexts between research and LMS submission.
While tools offer AI generation capabilities, there is currently no way to save insights directly without switching between tabs.
Ideation, Design, Iterate, Test, and Repeat
IDEATION
Initial Feature Exploration
Given the constraint of time, I sketched out many low-fidelity concepts to visualize how to integrate the feature into Perplexity.
USABILITY TESTING
Discovering New Perspectives and Iterating
Through 5 rounds of designing and iterating, I interviewed a diverse group of users - including those who had never used Perplexity and experienced Perplexity users - to ensure the design is intuitive for both new and returning users.
Key Change #1: Adding to New Spaces
90% of users preferred the inline dropdown over a popup for its speed. It lets them highlight, save, and keep reading without interruption, while also adding to a new Space without leaving the page.
Key Change #2: How Quotes are Displayed
I explored different visual styles for how saved quotes would appear. Through user testing, I found that users preferred a cleaner look.
FINAL PROTOTYPE
Snippet: Save, Organize, and Reuse Quotes Without Losing Context
Add Article Quotes to New Folders in Spaces
Select an article and highlight important texts and create a new Spaces folder and store the texts.
Write Essays with Saved Quotes
Type into an essay topic and generate essays with quotes embedded directly in your Spaces folder.
Use Snippet to Create Flashcards
Type in the text and Perplexity will generate you flashcards.
Final Testing Metrics
NEXT STEPS
If I had more time…
I would:
TAKEAWAYS
Good design reduces cognitive load.
Users are often juggling multiple tools and tasks at the same time — my job as a designer is to make the next action frictionless. When flow is uninterrupted, engagement follows naturally.
Cross-functional teamwork and collaboration drive product clarity.
Working closely with PMs and PMMs helped align design decisions with product goals, ensuring the every feature we created wasn’t just functional — but also launched with the right purpose.
Test. Learn. Iterate. Repeat.
Through rounds of usability testing and talking to users, I saw firsthand how tiny interaction tweaks — like confirmation states or button placement — made the difference between hesitation and intuitive use.