Designing a
Mobile App
for Smarter,
On-the-Go Learning
Table of Contents
1
2
3
4
5
6
7
8
Executive Summary
ROLE
UX & Product Lead
TEAM
1 UX & Product Lead (Me)
1 Assistant UX Researcher
1 Software Engineer
COMPANY
LearningClues, Inc.
OUTCOMES
90% student satisfaction rate
95+ design system assets
35 hi-fi screens in light/dark
TOOLS
Figma, Mockups, Miro, Jira
DURATION
3 Months
February - April 2025
CLIENTS
4
of the Top 50 Universities in the US
ACTIVE COURSES
30
Using the LearningClues System
STUDENT USERS
2500
?
A B2B AI-edtech startup.
GOAL?
Disclaimer
To comply with my non-disclosure agreement, I have omitted, obfuscated, or blurred confidential information in this case study.
Problem Description
The (Business) Challenge
There is a low usage and retention rate of the LearningClues system by the end-users (students).
Business Goal
Maximize the number of users using our system by providing an accessible, mobile-friendly version that intelligently reminds users to study and hence boosts engagement
Constraints
⏳ Limited Time → 3 months to conduct both research and design in parallel
📋 Investors demand a mobile app → no flexibility to explore other solutions.
Going a step further to see the user’s perspective...�
UNDERLYING USER PROBLEM
My Role
I led the end-to-end design of the mobile app from research to developer handoff within 3 months
User and Market Research
6 weeks
Developer Handoff
1 week
Design
6 weeks (2 weeks overlapping with research)
Usability Testing
1 week
February 2025
April 2025
Handed off
35
Hi-fidelity fully-prototyped screens in light and dark modes
Developer Handoff
Laying some Groundwork
What is Expected from the Mobile App at a Minimum?
An AI Chatbot that answers student queries based on their course content
Ability to create custom quizzes and automatically get insights on concept-level proficiencies
🧠 Some Food for Thought...
But is a mobile learning experience the same as a Desktop/Laptop experience? How does one account for the small screen size and competing distractions on mobile?
Drafting My Project Plan
RESEARCH
DESIGN
Research
50+ Survey Responses • 5 Interviews • 5 Market Competitors’ Analysis
Mapping Research Objectives → Problems → Methods
Research
Step 1: Drafting User Personas
Who will be the users of the mobile app? Starting with some assumptions about the user to get a sense of direction for the research...
Meet Larry
An undergraduate stem student with a demanding schedule and heavy workload
GOALS
PAIN POINTS
NEEDS
MOTIVATIONS
Research
Step 2: Comparative Review of Competitor’s Products
What should we be doing based on what is already out there?
Analyzing a mix of edtech companies and social learning platforms to get a holistic view of the competitive landscape..
To understand the competitor’s product workflows and identify key features, I generated a 64x19 comparative matrix
Outlined key workflows in competitors apps via interaction maps (image blurred for confidentiality)
To help other team members read and understand the matrix and increase efficiency of future use...
A short description of each feature being analyzed
Star to indicate features that were exceptionally implemented by competitors
A column and key to indicate whether a feature is present, missing, unclear or somewhat inferrable
Key Takeaways from Competitive Analysis
Going beyond just replicating what competitors are doing...
What sets us apart? Where is our unique value proposition?
Research
Step 3: Launching a Survey (reviewed by Assistant UX Researcher)
To better understand our target audience and recruit participants for follow-up interviews.
Through 50+ student survey responses, I learnt...
Grades are the primary study motivator. Study depth matters after
Mobile is system-driven while Desktop is user-driven
Micro-Learning on Mobile is Real.
*findings-level stats have been hidden for confidentiality
Research
Step 4: Deeper Insights through Semi-structured Interviews
Talking to Potential Target Users to Uncover Jobs-to- be-Done
“The Messy Middle”
A challenge I ran into...
A lesson I learnt...
Organizing findings from 5 interviews into an affinity map with 500 notes, to uncover 4 key themes...
Theme 1
Study Habits Depend on Context and Energy
Theme 2
Tracking Progress is Essential to Staying on Track�
Theme 3
Study Tools are Needed that Support Focus and Cognitive Load
Theme 4
Support Systems Help Push Through Challenges
Reshaping the Problem
Problem (Old)�
UNDERLYING USER PROBLEM
Problem Redefined�
UNDERLYING USER PROBLEM
Evolving the User Persona
Meet Larry
An undergraduate medical student with a demanding schedule and heavy workload. He commutes to campus by public transport daily and often feels like he’s not using his in-between time productively.
GOALS
PAIN POINTS
NEEDS
Feedback that helps him track progress and feel in control
Timely reminders and bite-sized, study tasks that fit into short breaks
MOTIVATIONS
*Changes are marked in blue
Design
95+ Components of Design System • 35 High Fidelity Screens in Light & Dark
I asked...
Our Solution
Introducing the LearningClues Mobile App: Your 24/7 Companion to Better Grades and Learning Mastery
A personalized AI study companion built for focus, feedback, and quick study with a mobile app
on-the-go.
“I want to know what to focus on when I have 15 minutes”
“I need quick stats and nudges, I don’t have time to review everything”
“My phone is for quick learning moments and laptop for prolonged study”
Personalized Practice
Daily 5-minute challenges focused on topics you need to improve or revise
Tailored Feedback
Get instant feedback, hints and explanations cited back to your course content
Performance Tracking
View your strengths and weaknesses so you can focus on what’s important
Tying the research to a solution
You asked, we delivered...
Key Design Decisions
How I found the sweet spot in mobile learning to boost usage without overwhelming the student...
1
Feature
Ready-made, bite-sized, and Personalized Study Material
Why?
Minimizes decision fatigue (from deciding what to study) and supports micro-learning moments.
Daily 5-min practice suggested based on your performance in course concepts
Auto-generated prompts for AI coaching sessions based on what you are struggling with
2
1
Feature
Targeted Performance Stats and Accountability Metrics
Why?
Summary Stats to get an overall sense of how you’re performing
Breakdown of proficiency for each concept, and ability to filter out low, medium and high proficiency concepts to identify strengths and weaknesses
2
1
Feature
Tailored Feedback with course-specific citations
Why?
Mobile friendly version of the LearningClues AI Chat that provides citations to course content
Ability to search the web and across courses
Get course-specific feedback to test question answers
2
3
The Building Blocks
From early design workshops to sketches and low-fidelity mockups
Step 1: Brainstorming
Goal?
Engage the team to brainstorm mobile features based on existing Desktop functionality.
Outcome?
A preliminary list of features including mini practice tests that can be transported from Desktop to mobile.
Step 2: Sketches
Goal?
Sketch out initial wireframes based on competitive analysis findings.
Outcome?
10 high-level unique screen sketches that were shared with the admin team for feedback.
Step 3: Lo-fi Mockups
Goal?
Incorporating insights from user research to refine sketches into lo-fi mockups and better communicate ideas to the team.
Outcome?
10 low-fidelity mockups focused on high level features including mini-quizzes, hints, explanations and stats.
Step 4: Design System Creation
Challenge?
Existing design assets lacked consistency in the absence of a design system.
Outcome?
17 collections with 92 components fully documented
Usability Testing
10 usability tests
Usability Test Findings and updates...
Usability Test Findings and updates...
The term “refresh” used to indicate that a concept needs revision is confusing to users so it was replaced with “Needs revision”
In-text citations made more prominent and recognizable per users’requests
The formula for the residual sum of squares is given by the sum of the squared differences between the actual y value of a training point and its predicted y value using the linear model. This is represented by the formula:
RSS = ∑ni=1 (yi - y^i)2
Where yi is the actual target value and y^i is the predicted value using the linear model with parameters W and B .
Citations:
Outcomes
80%
students completed all 5 assigned tasks successfully
8/10
said they would use it in their study routine
9/10
said they would recommend it to other students
Results
Is it usable and will students find it helpful?
Lessons Learnt
Problems can evolve. It’s important to be mindful and flexible of this change and adapt your design strategies accordingly.
You always have less time than you think you do. Plan ahead for last minute surprises like user recruitment challenges and leave some time cushion for fallbacks.
Lessons Learnt
Some of my key takeaways...
THANK YOU!
I’m excited to see the app launch and share real-time numbers about usage and engagement. Stay tuned!