SoulMate AI chatbot
A compassionate conversational AI offering 24/7 emotional support, breathing exercises, and grounding techniques, always one tap away.
Overview
The problem we set out to solve
Mental health is widely recognized as critical to overall well-being, yet accessible and personalized care remains out of reach for most people. We asked what a digital system looks like when it genuinely meets people where they are.
264M
people worldwide affected by depression (WHO)1 in 13
individuals experience anxiety disorders globally89.1%
of our survey respondents had never used a mental health app76.1%
preferred self-assessment before seeing a professionalThe gap
High costs, professional shortages, long wait times, and geographic barriers leave millions without the mental health support they need.
The shortfall
Existing apps like Headspace, Calm, and Mindspa each cover a slice of user needs. None offer a comprehensive, AI-driven, therapist-connected platform.
Our answer
SoulSupport combines AI assistance, real therapist booking, mood tracking, guided exercises, and mental health resources in one place.
Design goals
- Make mental health support geographically and financially accessible to underserved users
- Deliver AI-personalized plans based on individual wellbeing assessments
- Connect patients directly with verified, bookable therapists
- Reduce stigma through private journaling, a compassionate chatbot, and community features
- Protect sensitive user data with encryption and GDPR/HIPAA-aligned practices
Research & discovery
Understanding the user before drawing a single screen
A mixed-method phase: an interview with a licensed psychotherapist and a quantitative questionnaire aimed at a diverse demographic.
Expert interview
"Offline sessions are more effective for diagnosis. Body language, facial expressions, and environmental tools are irreplaceable. But online sessions are a breakthrough in accessibility. Reach patients anywhere in the world, at any time."
Dr. Mohamed Shokrey AboelSadat, Cognitive Behavioral Therapist
The interview surfaced three design principles: transparent pricing and scheduling, lightweight pre-session tests that supplement real interaction without replacing it, and exercises that are tailored rather than one-size-fits-all.
Survey findings · 46 respondents
Features respondents valued
% of respondentsHow people prefer to communicate
% of respondentsFace to face · 73.9%
Online · 26.1%
Most respondents want a person in the room, but a quarter are comfortable going fully online. That split points to a hybrid model.
Top challenges
Anxiety, stress, and depression dominated responses. Overthinking, burnout, and insomnia were notable secondary mentions across age groups.
Barriers to app use
Privacy concerns led the list, followed by cost, perceived usefulness, and technical friction. Trust is the foundation of this product.
Competitive analysis
We audited 6+ existing solutions. Select the features you care about to see which products cover all of them.
Select one or more features to compare coverage.
User personas
Who we are designing for
Two primary personas emerged from the research: the student in academic crisis and the professional managing chronic stress.
S
Sara, 21
University student · Alexandria
Mood trackerWellbeing testsSoulMate
- Experiencing academic burnout, social anxiety, and sleep disruption ahead of final exams
- Uncomfortable discussing mental health openly because of cultural stigma at home
- Wants a private, non-judgmental space to track mood and access self-care tools
- Can't afford private therapy and needs free or affordable initial support
GoalUnderstand what she's feeling and know when to seek professional help.
K
Khaled, 34
IT project manager · Cairo
Therapist bookingPersonalized planMood insights
- High-pressure role with tight deadlines, managing remote teams across time zones
- Experiences persistent stress and early signs of burnout but dismisses them as "normal"
- Tried a therapist before but found scheduling and cost prohibitive long-term
- Wants data-driven insight into his stress patterns and on-demand support
GoalBuild sustainable habits and reach a therapist when needed, on his schedule.
Ideation & architecture
Mapping the emotional journey and system logic
Before wireframes, we mapped the user's emotional arc and the branching logic that separates a routine daily check-in from an urgent de-escalation path.
User journey map
The curve is illustrative: it shows direction of feeling across stages, not measured values.
Dual app flow logic
Feature scope prioritization
Online therapy sessionsWellbeing testsMood trackingMental health resourcesPersonalized AI plan
Conversational AI (SoulMate)Therapist chatCommunity discussionEvents calendar
Self-care remindersGamification badgesVR sessions (future roadmap)
Design execution
From wireframes to a visual language of trust
The visual identity had one mandate: safety. Every color, type choice, and spatial decision was judged against a single question. Does this feel like a sanctuary?
Color system
Tap a swatch to copy its hex value.
Sampled from the final screens. The deep blue carries the brand and the primary actions. The pale blue canvas and surface tints keep large areas quiet. Warm peach and blush only appear inside illustrations and goal cards, so the warmth sits in the content and the controls stay blue.
Typography system
Display / hero
Healing starts here
DM Serif Display · 32pxSection heading
How are you feeling?
DM Serif Display · 22pxUI / body
Your mood has been tracked successfully. Check your weekly report.
DM Sans · 15px · 400Label / caption
Mood logged · 3 minutes ago
DM Sans · 12px · caps · 500DM Serif Display is editorial and warm; DM Sans is clean and accessible. Together they feel emotionally warm and medically trustworthy, which is unusual in a space that tends toward either clinical coldness or saccharine cheerfulness.
The product
The final screens, one by one
All twelve screens as designed, grouped by where they sit in the journey. Choose a screen, then tap the numbered markers to see what each part of the interface is doing.
Home screen rationale
The home screen follows one principle: the most important action must take the fewest taps. The SoulMate button sits in the top third of the screen, ahead of the shortcuts, test card and doctors below it.
Testing & iteration
What users told us when we got out of the way
We ran usability sessions on four core flows: onboarding, mood logging, chatbot interaction, and therapist booking. Here is what we learned and how we responded.
Key findings
Issue
Onboarding friction
Users were confused by the post-signup test screen. Many thought the assessment was mandatory and felt pressured, not welcomed.
Issue
Chatbot cold start
A blank chat state caused hesitation. Users didn't know what the AI could and couldn't do, which created anxiety rather than relief.
Issue
Therapist trust signals
Users wanted to see more before booking: credentials, session format (video/audio), and estimated wait time.
Insight
Mood emoji resonance
The animated emoji in mood logging had the highest delight score across sessions. Users felt it "actually understood" them.
Insight
Plan page underused
The personalized plan was buried. Users who found it loved it, but most missed it on first sessions. A discoverability failure.
Fix
Test skippability
Adding "Skip for now" to the initial assessment improved onboarding completion in the second round of testing.
Key design iterations
Results & impact
What SoulSupport delivers
The final product is a cohesive mental health ecosystem where patients, therapists, and administrators work within a single, secure, AI-enhanced platform.
12
core features built, covering AI, therapy, self-care, and community3
user roles fully designed: patient, therapist, and admin93%
of test users found mood tracking "genuinely useful" in round 22
rounds of usability testing, 6 major design iterationsFinal feature set
Therapist booking
A verified directory with rich profiles, real-time availability, video/audio/chat options, and integrated payment processing.
Mood tracking
Daily emoji-based logging with optional journaling, AI-detected patterns, and progress charts shared (with consent) with connected therapists.
Personalized AI plan
Post-assessment wellness programs combining meditation, exercise videos, journaling prompts, and curated reading.
Wellbeing assessments
Validated psychometric tests for anxiety, depression, life satisfaction, and self-compassion, framed to empower and never alarm.
Mental health library
Clinically reviewed documentation on mental health conditions, organized alphabetically and updated by the admin team against global statistics.
Designer's reflection
This project taught me that designing for mental health is an act of trust architecture. Every decision, from the warmth of a color to the phrasing of an error message, either builds or erodes the user's willingness to be vulnerable.
The biggest tension was between clinical credibility and emotional warmth. Too sterile and users feel surveilled. Too casual and they doubt the medical validity. The middle path is a product that feels like a conversation with a wise, caring friend who also happens to be a professional.
The next frontier is crisis detection: using AI to identify linguistic patterns in chat that signal acute distress, and surfacing emergency resources proactively instead of waiting for the user to ask.
What I'd do differently
- Recruit participants with lived mental health experience earlier. Student demographics alone don't capture the full user spectrum.
- Prototype the AI chatbot interaction earlier. Its emotional register needs more testing cycles than static UI.
- Co-design with a licensed therapist as a regular design partner, not only an interview subject.
- Run accessibility audits from the start. Mental health users often include people with anxiety, ADHD, or dyslexia who are underrepresented in standard UX research.