MediMind AI
MediMind AI
AI-powered Medication Adherence App
AI-powered Medication Adherence App
Problem
How many among us have forgotten to take our medication course as prescribed by our primary care physician? How many people do you know with chronic conditions that forget to or are unable to properly maintain their medication adherence due to lifestyle factors?
Current medication reminder apps only provide static reminders at fixed times. We often find ourselves too busy or too stressed to pay attention to or care about such notifications.
The app, MediMind AI, will go beyond reminders to become an intelligent health companion. It will learn from user behavior, health history, and adherence patterns to proactively support medication management, side-effect tracking, and habit formation — powered by personalized AI coaching.
Methodology
Defined user personas and informed survey design.
Evaluated UI/UX, core features, and business models of current meal planning and meal prep apps.
Conducted online surveys and interview and gathered quantitative insights from target users.
Methodology
Defined user personas and informed survey design.
Evaluated UI/UX, core features, and business models of current meal planning and meal prep apps.
Conducted online surveys and interview and gathered quantitative insights from target users.
Methodology
Defined user personas and informed survey design.
Evaluated UI/UX, core features, and business models of current meal planning and meal prep apps.
Conducted online surveys and interview and gathered quantitative insights from target users.
UX Research
To gain a better understanding of the pain points experienced by both patients and doctors, I conducted a series of surveys.
Patient Research
The patient survey had 12 questions to learn how people navigated their daily medications.
Doctor Research
Simultaneously, I also conducted a survey and an in-depth interview of a doctor to understand what they would and wouldn’t appreciate as part of their check-in process. The survey was 16 questions, and the follow-up interview ran for 15 minutes.
How often do you forget to take your meds?
What usually causes you to miss a dose?
Would you feel comfortable receiving health tips or medication insights from an AI assistant?
UX Research
To gain a better understanding of the pain points experienced by both patients and doctors, I conducted a series of surveys.
Patient Research
The patient survey had 12 questions to learn how people navigated their daily medications.
Doctor Research
Simultaneously, I also conducted a survey and an in-depth interview of a doctor to understand what they would and wouldn’t appreciate as part of their check-in process. The survey was 16 questions, and the follow-up interview ran for 15 minutes.
How often do you forget to take your meds?
What usually causes you to miss a dose?
Would you feel comfortable receiving health tips or medication insights from an AI assistant?
UX Research
To gain a better understanding of the pain points experienced by both patients and doctors, I conducted a series of surveys.
Patient Research
The patient survey had 12 questions to learn how people navigated their daily medications.
Doctor Research
Simultaneously, I also conducted a survey and an in-depth interview of a doctor to understand what they would and wouldn’t appreciate as part of their check-in process. The survey was 16 questions, and the follow-up interview ran for 15 minutes.
How often do you forget to take your meds?
What usually causes you to miss a dose?
Would you feel comfortable receiving health tips or medication insights from an AI assistant?
MVP Objectives
Caregiver Mode
Summarizes trends for caregivers without breaching privacy (e.g., “User has missed 3 evening doses this week”), requires caregiver input
Pill Interaction Verifier
Using medically approved data, the AI can check pill interactions and advise against adverse drug combinations
Side Effect Journal
Users log symptoms; AI detects patterns and gives suggestions (e.g., “Consider taking med after eating”)
Adherence Prediction
Works in the background by analyzing data and predicts when a user is likely to miss a dose (e.g., weekends, mood dips) and adjusts reminder style/timing
Add Prescription
Two options will be available. One to manually enter dosages information, another that allows user to take a picture of their prescription pamphlet and auto-fill the data.
Automatic Refills
AI keeps track of refill and expiry dates (as entered) and orders refill from local pharmacy when user is running low, user will be prompted to complete the transaction
MVP Objectives
MVP Objectives
Caregiver Mode
Summarizes trends for caregivers without breaching privacy (e.g., “User has missed 3 evening doses this week”), requires caregiver input
Pill Interaction Verifier
Using medically approved data, the AI can check pill interactions and advise against adverse drug combinations
Side Effect Journal
Users log symptoms; AI detects patterns and gives suggestions (e.g., “Consider taking med after eating”)
Adherence Prediction
Works in the background by analyzing data and predicts when a user is likely to miss a dose (e.g., weekends, mood dips) and adjusts reminder style/timing
Add Prescription
Two options will be available. One to manually enter dosages information, another that allows user to take a picture of their prescription pamphlet and auto-fill the data.
Automatic Refills
AI keeps track of refill and expiry dates (as entered) and orders refill from local pharmacy when user is running low, user will be prompted to complete the transaction
MVP Objectives
MVP Objectives
Caregiver Mode
Summarizes trends for caregivers without breaching privacy (e.g., “User has missed 3 evening doses this week”), requires caregiver input
Pill Interaction Verifier
Using medically approved data, the AI can check pill interactions and advise against adverse drug combinations
List Generator
Users log symptoms; AI detects patterns and gives suggestions (e.g., “Consider taking med after eating”)
Smart Prediction
Predicts when a user is likely to miss a dose (e.g., weekends, mood dips) and adjusts reminder style/timing
Add Prescription
Two options will be available. One to manually enter dosages information, another that allows user to take a picture of their prescription pamphlet and auto-fill the data.
Automatic Refills
AI keeps track of refill and expiry dates (as entered) and orders refill from local pharmacy when user is running low, user will be prompted to complete the transaction
MVP Objectives
User Trust in AI Products
In order to ensure users can trust in the program, there are a few considerations that need to be applied. These primarily center around privacy, clarity and communicated functionality of the end product.
Quiet Assistant
The application should not present itself or its capabilities as super-human, but rather as a quiet assistant that runs in the background and assists only as needed. Think butler rather than supercomputer.
Respect Privacy
The data will need to be stored locally, on the device being used to run the application. Adhering to ethical guidelines and data protection policies will also be crucial to maintaining confidentiality.
Proper Guidance
Understanding the app's capabilities and limitations will prevent users from asking it to complete tasks outside its purview and inevitably becoming disappointed, thus leading to algorithm aversion.
Transparent Loop
Having a human-in-the-loop system to verify and correct the AI as it analyzes user data will rectify any misleading statements it makes while also training the AI to do better in the future.
Quiet Assistant
The application should not present itself or its capabilities as super-human, but rather as a quiet assistant that runs in the background and assists only as needed. Think butler rather than supercomputer.
Respect Privacy
The data will need to be stored locally, on the device being used to run the application. Adhering to ethical guidelines and data protection policies will also be crucial to maintaining confidentiality.
Proper Guidance
Understanding the app's capabilities and limitations will prevent users from asking it to complete tasks outside its purview and inevitably becoming disappointed, thus leading to algorithm aversion.
Transparent Loop
Having a human-in-the-loop system to verify and correct the AI as it analyzes user data will rectify any misleading statements it makes while also training the AI to do better in the future.
User Trust in AI Products
In order to ensure users can trust in the program, there are a few considerations that need to be applied. These primarily center around privacy, clarity and communicated functionality of the end product.
Quiet Assistant
The application should not present itself or its capabilities as super-human, but rather as a quiet assistant that runs in the background and assists only as needed. Think butler rather than supercomputer.
Respect Privacy
The data will need to be stored locally, on the device being used to run the application. Adhering to ethical guidelines and data protection policies will also be crucial to maintaining confidentiality.
Proper Guidance
Understanding the app's capabilities and limitations will prevent users from asking it to complete tasks outside its purview and inevitably becoming disappointed, thus leading to algorithm aversion.
Transparent Loop
Having a human-in-the-loop system to verify and correct the AI as it analyzes user data will rectify any misleading statements it makes while also training the AI to do better in the future.
Quiet Assistant
The application should not present itself or its capabilities as super-human, but rather as a quiet assistant that runs in the background and assists only as needed. Think butler rather than supercomputer.
Respect Privacy
The data will need to be stored locally, on the device being used to run the application. Adhering to ethical guidelines and data protection policies will also be crucial to maintaining confidentiality.
Proper Guidance
Understanding the app's capabilities and limitations will prevent users from asking it to complete tasks outside its purview and inevitably becoming disappointed, thus leading to algorithm aversion.
Transparent Loop
Having a human-in-the-loop system to verify and correct the AI as it analyzes user data will rectify any misleading statements it makes while also training the AI to do better in the future.
User Trust in AI Products
In order to ensure users can trust in the program, there are a few considerations that need to be applied. These primarily center around privacy, clarity and communicated functionality of the end product.
Quiet Assistant
The application should not present itself or its capabilities as super-human, but rather as a quiet assistant that runs in the background and assists only as needed. Think butler rather than supercomputer.
Respect Privacy
The data will need to be stored locally, on the device being used to run the application. Adhering to ethical guidelines and data protection policies will also be crucial to maintaining confidentiality.
Proper Guidance
Understanding the app's capabilities and limitations will prevent users from asking it to complete tasks outside its purview and inevitably becoming disappointed, thus leading to algorithm aversion.
Transparent Loop
Having a human-in-the-loop system to verify and correct the AI as it analyzes user data will rectify any misleading statements it makes while also training the AI to do better in the future.
Quiet Assistant
The application should not present itself or its capabilities as super-human, but rather as a quiet assistant that runs in the background and assists only as needed. Think butler rather than supercomputer.
Respect Privacy
The data will need to be stored locally, on the device being used to run the application. Adhering to ethical guidelines and data protection policies will also be crucial to maintaining confidentiality.
Proper Guidance
Understanding the app's capabilities and limitations will prevent users from asking it to complete tasks outside its purview and inevitably becoming disappointed, thus leading to algorithm aversion.
Transparent Loop
Having a human-in-the-loop system to verify and correct the AI as it analyzes user data will rectify any misleading statements it makes while also training the AI to do better in the future.
Wireframes
Wireframes
Low-fidelity wireframes were built to show some applicable features, both AI and regular, that would make up this app's MVP.
Low-fidelity wireframes were built to show some applicable features, both AI and regular, that would make up this app's MVP.
