CASE STUDY ─ PRODUCT & UX DESIGN

Backtrack AI

Accessible Scoliosis Scanning & Wellness For Adolescents

Overview

Timeline

June 2024 - Nov 2025

TeAM

0 → 1 Product Designer
Founder (team of 2)

Responsibilities

UX Research
UX/UI Mobile Design
Prototyping
Vibe Coding (ML)

Tools

Figma
Claude
ChatGPT
Google Cloud

The Story

“

“It’s sophomore year of high school. One of my close friends is playing basketball. All of a sudden, he collapses. Struggling to breathe, he was hospitalized. Luckily, it was just a minor lung issue. However, as the doctor wrapped up their reports, they found something more serious that would end my friend's basketball career. A spine with a 17 degree curvature. Diagnosed too late for proper correction. Hearing this news broke my heart, yet, prompted me to learn more about scoliosis. “

My Journal, August 2024

finding the root problem

Through research, I discovered that Scoliosis affects 5% of the global population. Of which, Adolescents are at the highest risk of rapid curve progression during growth spurts.

$400

Is the average cost of each scan

3 weeks

Is the minimum wait for an initial consultation

85%

of scoliosis patients are left undiagnosed

Our Focus

Recognising that the adolescent demographic typically possesses limited disposable income, we aimed to address these challenges by creating an accessible solution: a mobile app with AI scanning

evaluating existing solutions

I first downloaded and actively tested existing posture scanning apps to understand common UI and UX patterns, while analysing user reviews noting the gaps of the apps.

Insights

UI

  • Nurturing colour palettes (coastal colours: sea-foam medical blue, soft pink)

  • Balanced negative and positive space (spacious layout)

  • Typography is friendly and non-aggressive

  • Tone and language is conversational and simplistic in language

  • Playful and smooth micro-interactions with clickable noises empowers users with interesting experiences

UX

  • Onboarding is simplistic, yet comprehensive to make users feel safe

  • Core scanning task flow has thorough instructions

  • Instead of purely focusing on scanning, most apps have additional functionalities, such as exercise library and recommendation feature.

  • User control can be improved with more customisation options in exercises

  • Motivation mechanics and progressive tracking features are desired, but limited

Key Discovery - The limited experiences of rehab exercises:

Neither home exercise programs (HEPs) nor existing apps engage users effectively, leaving them demotivated and disconnected. Consequently, many are unsuccessful with recovery. Therefore, I decided to challenge myself to design an additional exercise feature, one that appeals to adolescents.

how can I bridge this gap?

To further understand effective wellness motivation mechanics for adolescents, I conducted secondary research with articles, papers, and reviews. Considering there are not many successful wellness apps with adolescents as primary users, I decided to split my research into 2 parts:

  1. Analysing popular wellness apps to understand effective motivation and interaction design for general users

  2. Filtering relevant patterns and potential motivational features with apps that have adolescents as their dominant user base (Social platforms, EdTech)

After synthesising insights from each part, I identified overlapping themes and concepts. Shortlisting these behavioural design patterns provides me a clear and data-driven roadmap for reference during future ideation. Ensuring that I design features that are familiar to the existing wellness landscape, while resonating with adolescents. 

TOP 3 Behavioural Insights

Emotional safety

Adolescents needs to be affirmatively encouraged

Product Familiarity

Adolescents are more likely to use products that feel immediately useful

Authenticity Balance

Adolescents are constantly protecting their social image, while striving for raw genuine engagement

user interviews

I conducted the majority of primary interviews with general adolescents to gain additional insights on preferences and needs, while understanding their emotional journey. I also interviewed several existing patients to understand the landscape of current scoliosis rehab better.

user persona

The primary users are adolescents who are non-patients, since they are the majority of the population. I synthesised all the gathered data to a user persona to highlight their preferences, pain points, and behaviours.

Julian, the gamer

"After tomorrow's exam, I can finally play Brawl all day!"

Julian is a high schooler who spends most of the day on his computer replying to DMs, feeling slight neck and back pain regularly — but paying little attention to it. Finals are three days away, he's been procrastinating on his studies, and he hasn't exercised in weeks.

Motivation

  • Esteem (Maslow Hierarchy) - cares about fitting in socially

  • Pressured to get good grades from parents

  • Constant comparison amongst peers

  • Black hat driven - driven by incentives rooted in urgency, obsession, scarcity, and fear.

Pains / Frustrations

  • Frequently dealing with posture and back pain

  • Lack of motivation and time to consistently exercise

  • Uncertain about severity of posture problem

user journey

The primary users are adolescents who are non-patients, since they are the majority of the population. I synthesised all the gathered data to a user persona to highlight their preferences, pain points, and behaviours.

Julian

“As a teenager, I want to have a pain-free and healthy body so I can enjoy myself with friends and pursue my dreams!”

Design focus

Using insights from desk research, affinity mapping, user interviews, I decided to prioritise several design features, selecting ideas that are high impact, low effort, focused on bringing personalised and engaging experiences for adolescents. 

1 - “Duolingo”-fied experience 👾

Intention: adolescent-familiar and minimalistic design which instantly prompts users to start exercising once entering the app (without a classic “dashboard” home screen); breaking down exercise sequence to small and manageable steps to ease exercise routine habit integration

2 - Dinosaur Mascot 🦖

Intention: Builds a playful atmosphere, while simultaneously guiding users step by step, preventing confusion and information overload. 

3 - AI Exercise Builder 🏃‍♀️

Intention: provides adolescents with the autonomy to choose between tailored routines that are effective, while building their own if they have personal intentions.

4 - Daily Login Streaks and Notifications ✅

Intention: most effective nudging mechanism (based on research), increase user retention and motivation.

Core flows

After the synthesis and design goals, I crafted a comprehensive user flow to understand the various paths users may take when navigating the screen, diverging 3 core flows: scanning, exercise, and new to Scoliosis.

Part of the Onboarding / Exercise Recommendation Flow

Low-fid prototype

I focused most of my initial prototyping on different variations on the onboarding → scanning → exercise flow, since that’s the main flow for most users’ intention (to scan and check their backs)

mid-fid prototype + usability testing

After consolidating the general design direction, I built a functional prototype on Figma and moderated usability testing with my friends who previously had the thought of getting a scoliosis scan. The goal with this testing is to evaluate the interaction flows and identify if I missed any edge cases to consider for my final design. 

pitching - local and international

Presenting at the local and international round of the International Data Science Olympiad (ISDOL) gave me a fresh batch of user feedback with other participating and visiting students. In addition to the UI/UX design, I also gained critical insights to our product’s feasibility and business model by pitching to AI experts, professionals and founders. This allows me to consider potential constraints as I design new features moving forward. 

Key observations from testing:

  • Onboarding and scanning flows were direct, easy to navigate and engaging

  • Gamified home screen motivated users to instantly start exercising right after scanning

  • Custom exercise builder could provide more guidance and suggestions

  • Many commented on the low friction design of the exercise, which made them worried that they would skip through the exercise

Design priority identified - Based on pain points, I will work on improving the exercise section for my high-fid designs.

Final Iteration & Design

To address the user’s frustration, I took inspiration from screen-blocking apps that I use daily and added a “re-engagement” popup to create friction with the aim to help users reconsider skipping the exercise.

Here are my final designs:

Home Screen

Exercise Builder

Scanning & Report

Profile

Workout

Re-engagement Popup

Exercise Recommendation

Login Streaks

Key learnings

AI as a Thought Partner

Feeding Mobbin layouts to Claude and Google Stitch is highly effective to generate low-fid wire frame layout ideas and test quick mockups.

Minimalism and Usability Tradeoff

At times, it’s better to add an extra “confirm” button when users are about to commit to something bigger (like an AI report generation), enabling them to consciously reinforce their decisions and gain greater control to personalise their experience