Overview
Earthora is a mobile health app for a market where a doctor's visit can mean hours of travel and a full day's wage. It's built around Dr. Alex, an AI health assistant, appointment booking, and symptom tracking for patients, backed by a doctor portal for remote patient management and an admin dashboard for analytics. Every screen was designed to load in under 2 seconds on a 3G connection, because that's the connection most of rural Nigeria actually has.
An AI-powered mobile health app addressing Nigeria's healthcare accessibility crisis: telemedicine, AI symptom tracking, real-time consultations, and a companion doctor portal.
The Challenge
Nigeria faces a severe healthcare accessibility crisis, especially in rural areas, where timely medical consultation is often unavailable, expensive, or delayed. Patients frequently resort to self-diagnosis, while doctors have no efficient way to manage patients remotely.
The design objective was straightforward to state and hard to build: a single, end-to-end mobile health experience that lets users monitor their health proactively, consult verified doctors remotely, and get personalized, AI-driven health insights, all inside one app that doesn't feel like three apps stitched together.
Business Goals
- 1Onboard 150+ verified doctors in the first quarter, enough to guarantee consultation availability
- 2Earn strong weekly retention by positioning Earthora as a health companion, not just a transactional booking app
- 3Cut appointment no-shows dramatically through reminder flows that are automated but still well designed
- 4Keep the platform fully usable on 3G connections and the low-end Android devices most common in rural Nigeria
User Needs
- 1Get trustworthy answers to health worries without an expensive or distant clinic visit
- 2Understand symptoms in plain language, no medical jargon, no fear-inducing ambiguity
- 3Book, reschedule, and attend a consultation with a verified doctor in just a few taps
- 4For doctors: manage remote patients, follow-ups, and consultation notes in one efficient portal
Key Findings
- 1Patients feared misdiagnosis above everything else, whether from their own googling or from a rushed clinic visit
- 2Users wanted a health companion, not a transactional app. Check-ins and tracking mattered to them nearly as much as consultations did
- 3Doctors needed streamlined follow-up tools, since post-consultation care essentially collapsed the moment a patient left the clinic
- 4Observing clinic queues in Lagos showed patients routinely waiting 3+ hours for a consultation that lasted minutes
- 5Data cost and connectivity anxiety shaped behavior directly. Users avoided any app that felt heavy or data-hungry
User Voices
"I googled my symptoms and convinced myself it was something terrible. I could not sleep until I saw a doctor two weeks later."
"The clinic is far and the queue is long. Unless it is serious, you just buy something from the chemist and hope."
"After patients leave my consulting room, I have no way to know if they took the drugs or got worse."
User Personas
Affinity Mapping · Key Themes
Fear & Trust
- 1Fear of misdiagnosis drives both over-anxiety and dangerous avoidance, sometimes in the same person
- 2Verified-doctor badges and credentials are decisive trust signals
- 3AI advice only gets accepted when a path to a human doctor sits right beside it
Access Economics
- 1A clinic visit costs travel, queue time, and often a full day's income
- 2Data cost is a real adoption barrier: an app perceived as heavy simply doesn't get installed
- 3Telemedicine has to be visibly cheaper than the trip it's replacing
Companion, Not Transaction
- 1Users want ongoing check-ins, tracking, and reassurance in between illnesses, not just during them
- 2Weekly engagement comes from tracking streaks and insights that feel personal
- 3A companion tone in plain language beats clinical detachment every time
The Broken Follow-up Loop
- 1Doctors lose all visibility into a patient the moment the consultation ends
- 2Patients forget instructions and drop off medication partway through
- 3Structured follow-ups and reminders close that loop for both sides
Empathy Map · Amina Bello
User Journey Map
Information Architecture
Patient App
- 1Dr. Alex AI assistant
- 2Symptom tracker
- 3Appointment booking
- 4Teleconsultation room
- 5Health records
Doctor Portal
- 1Patient management
- 2Consultation schedule
- 3Video consultation
- 4Notes & e-prescriptions
- 5Follow-up tracking
Admin Dashboard
- 1Doctor verification
- 2Platform analytics
- 3Content moderation
- 4Support operations
Health Blog
- 1Editorial articles
- 2Condition guides
- 3Personalized feed
Shared
- 1Notifications & reminders
- 2Payments
- 3Messaging
- 4Settings & privacy
User Stories
Approaches Considered
Telemedicine Marketplace
A focused doctor-booking product: search verified doctors, book a video consultation, pay. No AI layer, no tracking.
Pros
- + The simplest scope, and the fastest to launch
- + Clear revenue per consultation
Cons
- - Purely transactional, with no reason to return between illnesses
- - Ignores the self-diagnosis behavior driving the whole crisis
- - Leaves the doctor follow-up loop just as broken
Connected Health Ecosystem
A companion-first ecosystem where Dr. Alex's AI triage and symptom tracking feed into teleconsultations, a doctor portal closes the follow-up loop, and an admin layer governs quality.
Pros
- + Meets users right where the journey starts, at the symptom scare
- + Companion features drive retention well beyond individual illness episodes
- + The doctor portal closes an adherence loop no competitor was addressing
Cons
- - The largest scope of the three, spanning four components
- - AI triage demands careful safety and trust design
AI Symptom Checker
A standalone Dr. Alex triage app that refers users out to partner clinics instead of hosting consultations itself.
Pros
- + A light footprint, ideal for low-data contexts
- + Real potential to spread virally
Cons
- - Referring users out loses them at the exact moment they need the product most
- - No doctor-side product means no defensible ecosystem
- - Weak monetization without consultations in the mix
Design Approach
Every design decision passed through two filters: does it build trust, and will it survive a 3G connection on a low-end Android phone? The visual language is calm and clinical-warm, soft depth, generous type, almost no alarm-red except where it's clinically warranted. I deliberately designed Dr. Alex as a companion rather than an oracle: plain-language responses, explicit confidence framing, and a consult-a-doctor path rendered alongside every single AI answer. The doctor portal borrows density patterns from professional tools while keeping the consultation flow itself distraction-free. Performance was treated as a design constraint, not an engineering afterthought: image budgets, skeleton states, and offline-tolerant flows were all specified directly in Figma.
A 60+ component design system spanning patient, doctor, and admin surfaces, with accessibility, low-bandwidth states, and light and dark variants built into every single one.
Key Design Decisions
- 1
Dr. Alex always pairs AI guidance with a visible consult-a-doctor path, so the AI complements human care and never dead-ends a worried user
- 2
Plain language as a system rule: every clinical term gets a one-line explanation, tested directly against non-medical readers
- 3
A three-step booking flow (doctor, time, pay), cut down from the original seven after journey mapping exposed serious drop-off risk
- 4
Verified-doctor credentials, photos, and consultation counts surfaced at every decision point to push back against deep telemedicine skepticism
- 5
A sub-2-second load on 3G as a hard design budget, which drove skeleton screens, aggressive image discipline, and offline-tolerant symptom logging
Patient Mobile App
The patient app is the heart of Earthora, built to feel like a health companion rather than a transactional tool. Users talk to Dr. Alex, the AI health assistant, for instant symptom analysis and personalized health insights. From there they book appointments with verified doctors, track health metrics, browse curated wellness content, and manage their full medical profile, all of it designed to work on low-bandwidth 3G connections across Nigeria.
Doctor Mobile App
The doctor-facing app makes remote patient management workable on the go. Verified doctors manage their patient roster, run teleconsultations over secure video, review AI-generated patient summaries, and schedule follow-ups from a clean, clinical interface built for mobile. The calendar system, paired with automated reminders, cut appointment no-shows by 90% during testing.
Landing Page & Health Blog
The public-facing website carries Earthora's mission to bridge Nigeria's healthcare gap. A clean, trust-driven landing page turns visitors into app users through clear value propositions and social proof, while the integrated health blog delivers curated wellness content, everything from nutrition guides to mental health resources, building Earthora's credibility as a health authority over time.
Method
Task-based user acceptance testing with A/B follow-ups
Participants
50+ users across patient and doctor roles
Task Success
89%
Test Tasks
- 1Ask Dr. Alex about a symptom and interpret the guidance
- 2Book a video consultation with a verified doctor
- 3Log allergies and an ongoing condition into your health profile
- 4As a doctor, review a patient's symptom history and record consultation notes
- 5Set up medication reminders from a care plan
Key Findings
- 1Allergy and condition logging confused non-medical users. The form assumed clinical vocabulary, so I rebuilt it around plain-language selection instead
- 2Symptom checker results needed clarity work. Users wanted explicit next-step guidance, not just a probability-style output
- 3A/B testing on appointment reminders showed push notifications decisively beat in-app badges at reducing no-shows
- 4The consult-a-doctor path sitting beside AI answers measurably raised trust scores in post-task surveys
- 5Once note templates were introduced, doctors completed consultation documentation faster than their old paper baseline
01 · Booking flow was hemorrhaging intent
Before
The original appointment flow ran seven steps long, and journey mapping plus testing both showed users abandoning it partway through.
After
I compressed it to three steps (doctor, time, payment) with progressive disclosure for optional details. Completion rates rose sharply, and the pattern shipped to production.
02 · Symptom checker outputs read as ambiguous
Before
Results presented a list of possible conditions with no clear severity framing, leaving non-medical users unsure what to actually do next.
After
I redesigned results around plain-language severity bands and a single recommended next action, with consult-a-doctor always sitting one tap away.
03 · Reminder strategy was underperforming
Before
Appointment reminders relied on in-app badges, and users simply weren't seeing them in time.
After
I shifted to a push-notification-led reminder sequence, validated through A/B testing. That change is what underpins the 90% reduction in no-shows.
Weekly Return Rate
Appt No-Shows Reduced
Doctors Onboarded (3mo)
Patient Satisfaction
Outcome
Earthora became a trusted health partner for thousands of Nigerians: a 75% weekly return rate, and a 90% reduction in appointment no-shows driven by automated reminders. 200+ verified doctors came onboard within 3 months of launch, and patient satisfaction landed at 4.8/5. The app loads in under 2 seconds on 3G connections across rural Nigeria, which is what makes quality healthcare genuinely reachable for underserved communities.
What We Learned
Healthcare design demands real clarity and real empathy. Medical jargon creates anxiety, and plain language is what builds trust back.
Localization matters at every level, from data usage constraints to cultural attitudes toward telemedicine in Nigerian communities specifically.
An AI health assistant has to complement human doctors, never replace them. The design always kept a consult-a-doctor path visible alongside every AI response.
