MedIntMMI
Customer Journey Design
Designing the end-to-end customer experience for an AI-powered medical MMI interview preparation platform — from onboarding through to performance benchmarking.
Role
Product Designer
Timeline
Dec 2025 – Present
Location
Hong Kong
Tools
Figma · Google Stitch · Miro
The Challenge
Medical MMI (Multiple Mini Interview) preparation is notoriously complex — candidates must develop a broad range of competencies across ethical reasoning, clinical scenarios, communication, and time pressure.
MedIntMMI needed an app experience that could guide different types of users through personalised learning journeys, without feeling overwhelming or clinical.
The Approach
I designed four distinct customer journeys — each mapped to a different user persona and learning mode — using Google Stitch and Figma.
Each journey was designed from scratch: user research, flow mapping in Miro, wireframing, then high-fidelity UI screens with a cohesive dark-mode design system built around depth, clarity, and momentum.
The Four Customer Journeys
Each journey maps a distinct user mode — from habit-building daily practice through to full exam simulation.
Onboarding to First Practice
A structured 5-screen onboarding flow that personalises the experience before a single question is attempted — collecting interview dates from 42 UK schools, experience level, and name. The user lands on WelcomeCard and completes their first AI-evaluated answer.
42 UK medical school picker with interview date entry to personalise the study plan
Daily Challenge CTA on WelcomeCard drives immediate first action (streak 0, sessions 0)
AI Evaluation result with score ring (7.5/10), dimension scores, and STAR + Four Pillars framework detection
Targeted Smart Practice
A returning user targets their weakest category using Smart Practice — a weighted random selector that surfaces the lowest-scoring category automatically. Records audio and video, receives AI feedback with framework detection, and shares a qualifying answer to the community.
Smart Practice tile auto-selects Ethics & Professionalism as the weakest category (avg 5.4/10)
CARE Framework + Four Pillars detected in AI evaluation — score 8/10 with timing card
Community share unlocked at 7+ threshold — answer anonymised before submission
Mock MMI Circuit
A full multi-station circuit simulation — from the MockCircuitSetupView (question count, time limit, difficulty, category toggles) through per-station recording with countdown timers and bell notifications, to aggregate results and percentile benchmarking.
Circuit setup with stepper (6 questions), time picker (25 min), difficulty (Medium + Hard), and category toggles
Aggregate score ring (7.4/10) with per-dimension breakdown: Structure 7.2, Communication 8.1, Reflection 6.8
Percentile benchmarking ranks the user Top 23% against all circuit completions
University Research & Interview Prep
From browsing 42 UK medical school guides (filtered by MMI format, region, and UCAT score) to reviewing a UCL school detail view, adding an interview date, and building mastery through NHS Hot Topics and spaced-repetition flashcards using the SM-2 algorithm.
Uni Guides filter by MMI format, England region, and UCAT 1800-2100 — 8 schools matched
Interview countdown triggers a 6-week adaptive study plan (47 days to UCL interview)
SM-2 flashcards with Again / Hard / Good / Easy rating — Ethics Principles deck, 4 cards due
Design Outcomes
Each mapped to a specific persona and learning mode, from habit-building to exam simulation.
The Obsidian Navigator system — built for depth, editorial clarity, and dark-mode legibility.
From user research and Miro flows through to high-fidelity Figma screens and Stitch prototypes.
Interested in this project?
Let's talk about how I can bring this kind of thinking to your product.