Lumi
An emotion-tracking app designed to help users build emotional literacy and discover supportive next steps.
Project
Mobile Web App (Course Project)
Role
Entire product design from research to conception, visualization and testing
Timeframe
6 months
About the project
In today’s fast-paced world, many people struggle to identify and manage their emotions leading to stress and overwhelm. Lumi started with a simple question: how do you help people name what they're feeling — and figure out what to do next — without making it feel like work?
I explored this as part of CareerFoundry's UX/UI Design program, working through the full design process independently. Competitive analysis revealed that existing emotion-tracking apps tend to fall into two camps: quick mood check-ins that stay surface-level, or deeper exploration that can feel complex and lacks clear guidance. This gap shaped Lumi's direction: a calm, guided experience that balances low effort with meaningful insight.
From user interviews through wireframing, prototyping, and usability testing, I grounded the tracking flow in therapeutic frameworks, developed a visual system tied to emotion categories, and tested and refined the experience with real users.
From research to personas
After competitive analysis revealed the gap in existing apps, I wanted to understand how people actually identify and manage emotions—and where they need support. I ran an online survey with 19 participants and 4 in-depth interviews.
Two user patterns, one shared core flow
The biggest surprise: emotional literacy varies widely. Some struggle to name emotions beyond "good" or "bad," while others are emotionally skilled but want support building consistency and deepening their practice. Using affinity and empathy mapping, I translated these insights into two personas representing these distinct patterns. Both personas use the same core features — but with different priorities and motivations.
Building the structure
Defining important tasks
Based on the personas and research, I identified three core tasks and mapped them as user flows: add an emotion entry, view insights, and do an exercise or read an article.
The emotion entry flow became the foundation—it enables both pattern recognition and emotion-linked support. I designed a two-step tracking approach that guides users from general emotions to more specific ones, with optional context to support meaningful insights over time.
Creating and testing the sitemap
I translated the flows into a sitemap and validated it through two rounds of card sorting. Key outcomes: separating Exercises from Articles, and choosing more approachable labels—like "Explore" for articles and "My Emotions" for insights and patterns.
From Wireframes to Prototype
With the structure in place, I used rapid sketching to explore design and navigation patterns through paper wireframes. Moving to digital wireframes helped me refine visual hierarchy and focus on essential elements, balancing clarity and guidance while maintaining a calm, uncluttered interface. Finally, I built an interactive prototype for usability testing.
Shaping the emotion-tracking feature
I based the emotion tracking feature on Stavemann's Feeling Star—a framework from behavioural therapy. I translated the model into a flow that guides users from core emotions to nuanced intensity levels, helping them name what they're feeling more precisely.
Putting the solution to the test
To validate the core flows and interactions, I conducted moderated usability tests with 6 participants using a mid-fidelity prototype. I synthesized findings in a Rainbow Spreadsheet and used Nielsen's severity scale to prioritize issues based on how users navigated the app, completed core tasks, and understood the tracking flow.
Understanding what works and what does not
Participants responded positively to the tone and navigated easily overall. However, testing revealed friction around gestures, navigation paths, and clarity—leading to improvements in the tracking flow, confirmation screen options, and copy.
Establishing Visual Design
Developing a visual direction
After validating Lumi's core flows with a grayscale prototype, I wanted to ensure the visual design supported emotional clarity without adding overwhelm. I ran a preference test comparing two approaches. Most participants preferred the Soft direction, describing it as calmer and more approachable—so it became the foundation for the final UI.
From direction to system: One emotion = one color
Lumi's core features—emotion tracking, My Emotions, and emotion-linked exercises and articles—work as one connected system. A consistent visual language helps users move between features without re-orienting, reinforcing emotional continuity rather than breaking it.
The Final Solution
Core flow 1: Adding an emotion entry
Tracking stays quick and low-effort while supporting deeper reflection. Optional context focuses on situational triggers and body sensations—grounded in research and emotion theory to generate meaningful insights.
Core flow 2: Reviewing patterns in My Emotions
My Emotions turns daily tracking into practical insight. Users spot patterns through an overview of emotions, triggers, and body sensations or switch to the calendar to revisit specific entries.
Core flow 3: Doing an exercise
Exercises are organized into needs-based categories and mapped to emotions so users can quickly find relevant support. Clear previews reduce uncertainty, and saved exercises make it easy to return to what worked.
Reflection
This project taught me to balance depth with focus. Working independently meant managing my own priorities—I had to regularly check whether I was pursuing what would deliver the most value or getting lost in details. After user research, I prioritized features like pattern insights and emotion-linked exercises while keeping tracking minimal enough for daily use.
The research also challenged my assumptions. I'd expected most people would struggle to name emotions, but emotional literacy varies widely—and both beginners and experienced users shared the same core need for frictionless tools that deliver real insight. Exploring how one design could serve both groups shaped key decisions like the two-step emotion identification flow and emotion-linked navigation.
If I were to continue, I'd test whether users sustain the habit over time, add depth to My Emotions without cluttering the interface, and expand exercises based on what users find most valuable.