Bridging Medication Management and Digital Literacy with AI
time
5 months
Scope
Concept
User Experience Design
User Interface Design
Brand + Visual Design
Prototype
Tools
Figma
FigJam
Notion
Adobe CC
Overview
Simplifying Medication Management
MedTracker is a case study exploring a mobile app designed to simplify medication management for individuals across all levels of digital literacy.
This project creates an inclusive and effective medication management experience, in part aided by the integration of AI through a voice-assisted interface named Vita - a latin term associated with life, vitality, and well-being.
My personal experience with my parents' medication management challenges motivated me to start this project, emphasizing the importance of catering to users across the digital literacy spectrum.
Challenge
The Problem with Current Solutions
Medication management is critical to healthcare, yet many individuals struggle with adherence, organization, and accuracy. Existing digital solutions often fail to address the needs of users with varying levels of digital literacy, leaving a significant gap in accessibility and usability.
Witnessing the challenges of manual medication management motivated me to create a reliable and user-friendly solution. My goal was to design an interface that could emulate the simplicity of pen and paper while leveraging digital tracking advantages.
The challenge lies in creating a solution that:
Simplifies medication tracking for individuals with limited technical skills
Provides a reliable and intuitive interface for managing complex medication regimens
Addresses the common pain points of existing medication tracking apps (e.g. confusing interfaces)
Leverages the potential of AI to enhance user experience while ensuring data privacy and security
Some factors complicated the development of this solution:
Balancing Simplicity and Functionality: Providing sufficient features for power users without overwhelming novice users.
Integrating AI: The technical complexities and the need for specialized expertise.
Hypothesis
My Hypothesis for Improved Tracking
My initial hypothesis was that semi-digitally literate and novice users could effectively leverage a voice-assisted feature to take advantage of digital tracking and inventory management. This functionality would differentiate MedTracker from existing solutions. Additionally, I aimed to gain practical experience integrating industry-standard AI into a digital product.
Given the limitations of this project, I explored comfort levels with similar voice-enabled services, how they might be used, and potential features that could be leveraged with the voice assistant.
Process
Designing the Solution
The design and development of MedTracker followed a user-centered process that integrated various research methods.
Research
Insights from Digital and Physical Methods
Direct Research: Digital Medication Tracking Apps
I conducted a detailed analysis of six prominent medication tracking apps (shortlisted from 22 apps across Android and Apple app stores) to understand user preferences and pain points.
Key Takeaways:
Users prioritize ad-free experiences, flexible reminders, and comprehensive databases.
Common pain points include notification issues and confusing interfaces.
The in-depth app analysis helped me choose my basic feature set:
Profile creation
Adding appointments/healthcare professional information
Logging medicine and setting reminders
Logging and tracking inventory
Views for charts, calendars, and lists
Editable medication detail pages
Health report data
Indirect Research: Physical Medication Tracking Methods
I also explored physical methods to understand their limitations:
Paper/Pen: Simple but prone to errors.
Pre-Dosed Services: Convenient but inflexible for complex regimens; very rare and not easily accessible.
Physical Calendars/Checklists/Logbooks: Visual but lack automation.
Physical Pill Dispensers: Organize medications but lack reminders.
Smart Bottles/Caps/EAMDs: Offer automated tracking but can be expensive and confusing for some.
Research Synthesis and User Personas
My secondary research, exploration of physical products, and competitor analysis informed the development of user personas to guide design decisions:
AI Exploration
Finding MedTrackers’s Voice
My research centered around understanding voice interfaces as well as how user interactions were organized. Unlike apps like ChatGPT, Perplexity, and Google Gemini, which organize interactions into threads or history sections, I chose not to use this pattern for MedTracker. Since users would likely not need to separate conversations to track actions, this approach kept database management simpler and more centralized.
Key Takeaways:
Clear visual indicators are essential for guiding user interactions in a new medium.
Voice input should complement text input for accessibility.
AI interaction history does not require complex categorization.
Technical Implementation: Vita’s Blueprint - Concept and Core Functions
I found integrating the voice assistant the most challenging phase. As a non-technical expert, I aimed to understand the considerations required to integrate a voice assistant into an app. In a real-world scenario, I would rely heavily on a technical team or subject matter expert for an in-depth workflow.
I adopted a two-phase parallel approach using the Google Assistant SDK and Dialogflow, balancing speed, user feedback, and long-term differentiation. Once a custom AI integration is ready, transitioning from the Google Assistant SDK to the custom solution can enhance AI capabilities.
Building a functional prototype with the Google Assistant SDK helps gather user insights on core features, voice/text command usability, and overall user experience. This informs and motivates the development of the Dialogflow model on the backend using real-world user data.
For long-term scalability, this workflow can be used with additional AI services or APIs. As MedTracker evolves, there may be market potential for more sophisticated AI capabilities, such as personalized recommendations, expandable conversational flows, predictive analytics, or integration with healthcare services. The flexibility of a custom AI integration will be essential for achieving these goals, especially cross-platform compatibility.
Integrating AI APIs might be complex initially, requiring extra tools which separates voice recognition and text-to-speech processing for a seamless experience in MedTracker. For example, using ChatGPT involves:
Voice Input -> Speech-to-Text (Google Cloud) -> ChatGPT (OpenAI) -> Text-to-Speech (Google Cloud) -> Voice Output
I prioritize user privacy in data handling by starting with local storage and a limited medication database. Users can opt-in for database updates if needed.
On the front-end, these are my proposed purpose and functions for Vita. The assistant functions as a voice or text input control to allow users to perform actions like:
Adding medications
Managing inventory
Answering questions by referencing data provided by the user within the local database
Modifying entries
Analyzing photos/documents of prescriptions or medications
Reporting adherence analytics
Setting reminders
Adding medicine doses on demand
As MedTracker continues to evolve, its potential to integrate with broader healthcare services and enhance user experience through AI-driven features will be crucial. By focusing on simplicity, accessibility, and user-centered design, MedTracker can set a new standard in medication management apps.
Design
From Sketches to Screen
With a feature and function set selected, I began to shape MedTracker's interface keeping the ease of use for novice users a priority. The design prioritizes simple section organization, at-a-glance information gathering, and low degrees of separation between function and result. This, in theory, is made even easier by Vita's assistance in the app.
The visuals highlight the progression of ideas and refinements that shaped the final product.
User Testing
User testing with a semi-digitally literate and digitally literate individuals helped inform certain UI decisions during the design process. The more significant changes included -
Making navigation more prominent in the app by simplifying labels and increasing touch targets
Direct access was added for returning to the "More" page.
Keeping onboarding icons and presentation more consistent between voice and manual flows.
Attachment icon confusion prompted updates for clarity. Older users related better with the paperclip icon versus something more modern (eg: a paper with an upward directional arrow or a + symbol)
A dark mode concept was made for a few key screens.
Brand
Visualizing Medical Adherence
Considering the wide spectrum of users, the app's core values are centered around simplicity, privacy and safety. These values are showcased in the user interface, the writing, and the overall structure of the app.
This section showcases the key brand assets developed for MedTracker, illustrating how these elements contribute to the app's overall identity and user experience.
Logo
The MedTracker logo combines a pill, clock, and medical cross in a radial style representing medication adherence and healthcare.
Colour Palette & Typography
The colour palette was carefully selected to balance trustworthiness, calmness, and clear communication.
Blue Ribbon (#364ce4) was chosen as the primary colour to portray reliability and calmness, contrasting with the urgency of red typically associated with medical emergencies.
Mesveda and Nacelle were chosen for their easy readability and modern style for an accessible and engaging user experience
Motion
Animating User Experience
Dynamic motion explorations were conducted for Vita's different states (listening, thinking, responding, talking, and processing), the "typing on" animation for Vita's responses, the MedTracker logo reveal, progress trackers, notifications, and transitions in the Figma prototype. The goal was to provide intuitive visual feedback, guide users, and add a sense of character to Vita’s presence.
A learning about Figma is that it has finite capabilities for heavy animation—I would use dedicated prototyping tools for complex animations in the future.
Insights
Impact and Future Potential
By offering a simple, reliable, and personalized experience, MedTracker has the potential to improve medication adherence, reduce errors, and empower users to take control of their health across all digital literacy levels.
My key learnings include recognizing the need for novice users in existing apps while understanding how AI integrations require specialized knowledge.
Explorations
Several alternate directions were explored but ultimately abandoned in favour of streamlined focus:
An SOS feature deemed less essential as users would likely turn elsewhere in emergencies.
Creating mood/habit tracking features were avoided to maintain focus on core tasks.
A complete voice-only interface was impractical due to prototyping limitations and edge case considerations. This made user testing difficult to conduct in a viable way.
Business Potential
The project also has significant business potential:
Partnerships with pharmacies and healthcare providers: MedTracker could be integrated into existing healthcare ecosystems to provide added value to patients and improve medication adherence rates.
Subscription models: Premium features, such as advanced analytics, personalized insights, and integration with wearable devices, could be offered through subscription models.
Data analytics for research: Anonymized and aggregated data could be used for research purposes to identify trends in medication usage and adherence, providing valuable insights for healthcare providers and pharmaceutical companies.
Wearable device integration: The mutual analytical and device sync advantages would benefit both the users and the development potential of the app
I am currently assessing the viability of building this app by reaching out to developers, subject matter experts, and potential investors to gauge resource requirements and interest.
By prioritizing accessibility, simplicity, and personalization, MedTracker has the potential to make a real difference in the lives of individuals who struggle with medication management.
