kulesza.io project hero img

Medical app

Built a time-boxed proof of concept for a self-diagnosis assistant that turns symptoms + basic user data into safe next steps: OTC suggestions or a GP visit. The app would then suggest widely accessible medications or propose an appointment with the local GP.

Project snapshot

Timeframe

July 2020


48 hours, time-boxed PoC

Discipline

  • UX research
  • UX design
  • UI design

Deliverables

Clickable proof-of-concept prototype

MVP scope + feature prioritization

Key flows (self-check → suggested next step)

Survey insights + distilled needs

Toolchain

  • Figma
  • Google Surveys - quantitative research

Discovery

Initial challenge analysis: 15-Minute Brainstorm

Writing down initial thoughts (part of the brand sprint) to gain insight into the problem and explore solutions (using the golden circle rule).

Golden Circle framework diagram with three columns: Why (reasons the app should exist including saving patients money and reducing clinic queues), How (methods like easy diagnosis process and user testing), and What (solutions including mobile app and landing page).

Success metrics

Preparing the most important KPIs for the MVP version to measure if the solution aligns with user needs.

Six KPI cards for MVP measurement: App downloads, Closing app after diagnosis, App rating, User database, User engagement, and Session time—each with a brief description of what to track.

User personas and their behavior patterns

Based on the most common user personas identified during competitive analysis, I sketched a simple flow showing the typical ways users achieve their goals.

User journey flowchart showing three paths from experiencing health symptoms: using the self-diagnosis app (leading to diagnosis and recommendations), searching the web (risk of inaccurate diagnosis), or visiting a GP (potential time and money waste).

Competitive analysis

Analyzing the most popular competing apps by building user flows from screenshots, evaluating features, and identifying the most optimized workflow.

Comparative user flow analysis of four competing symptom checker apps: WebMD, Symptomate, Ada, and Healthily—showing screenshot sequences from onboarding through symptom input to diagnosis results.

Exploring health decision-making patterns

I conducted a brief survey to interview potential users. I surveyed 49 people between the ages of 20 and 50, representing diverse industries and lifestyles. The primary objective was to validate various features and user needs.

Survey results from 49 respondents displayed in pie charts and bar graphs: health decision-making behaviors, doctor visit anxiety (63.3% open to self-diagnosis apps), preferred symptom input methods (61.2% prefer writing), useful health data categories, and interest in GP contact feature (89.8% positive).

Ideation

Planning the MVP scope

Based on the previous exploration, I prepared an MVP scope that includes additional features for future releases.

Product roadmap with three release phases: MVP (onboarding, symptom selection, interview, diagnosis, geolocation), v1.0 (Android app, landing page, accounts, dashboard, history), and v2.0 (activity app integration, medicine discounts).

First sketch of user-flow

The main goal was to create a general overview using low-fidelity wireframes. These wireframes would show how the app should guide users through the three main scenarios for the proof of concept:

  • The user wants to check symptoms and understand what might be happening
  • The user wants to confirm whether a doctor visit is needed
  • The user wants to check symptoms and learn possible self-care options
Low-fidelity user flow diagram in Figma showing the complete app journey: onboarding questions, symptom search with autocomplete, symptom selection, interview process, and diagnosis results with multiple screen states and decision points.

Build

UX prototyping

Using the Ant Design library, I created high-fidelity wireframes, and prepared an interactive prototype in Figma. This allowed me to conduct initial usability tests with potential users.

High-fidelity interactive prototype in Figma built with Ant Design library, showing connected wireframes for the full symptom-check flow from account creation through diagnosis, with visible interaction links between screens.

Design the general UI

Designed the key UI screens to communicate the product vision and critical moments of the flow.

Final UI design for key app screens: symptom input with yellow tag chips and common symptoms, blue-themed search with autocomplete suggestions, search results with symptom descriptions, and diagnosis screen with color-coded urgency alerts and condition cards.