PRODUCT DESIGN
Eye Health Tracker (Smart Home + AI)
Using smarthome devices and AI to improve my eye health.
AI
Powered
Smart Home
Integration
API
Communication
Summary
Project: Personal
Industry: Wellness & Health
Timeframe: September 2026 - Present
Role: AI Product Designer
Tools: Claude, HTML\CSS, Wix Studio, OpenAI, and Switchbot devices
Problem
My eyes are easily irritated due to an imbalance in the protective layers of my tears (dry eye syndrome).
Different factors can trigger flare-ups. These factors are not static, and it is very difficult to track down what is causing the discomfort. Sometimes it’s the weather, and sometimes it’s simply too much humidity in the air.
Solution
I created an AI health tracking system that helps me discover what is currently affecting my eyes. The system analyzes three data sources: home environment, outdoor environment, and user feedback (myself).
How does it work?
The AI-health tracker has 3 main data sources:
01
Indoor data:
Collected by smart home devices such as thermostats, air and humidity sensors all provided by SwitchBot devices.
02
Outdoor data:
collected through a local weather API (Open-Meteo), in this case for Calgary, Canada.
03
User data:
I designed a simple questionnaire that serves as daily feedback. The questions are based on dry eye assessments from ophthalmologists.
How is everything connected?
This is the most interesting part of the project. SwitchBot allows developers to collect data through a secure API. This data is automatically stored in my personal CMS server alongside environmental data and the feedback I provide daily.
What is the role of AI?
AI has been trained to look for patterns and trends. For example, if my apartment has been dry and cold and I, the user, then report irritation, the AI flags it and suggests logical recommendations. The AI is trained to analyze 15 days of data, which allows the system to find reliable patterns.
How is screen time accurately measured?
I have set up a motion sensor tracker right in front of my monitor so it can track the amount of time I’m actually spending in front of my screen. More sensors could be placed, but this setup works well for my routine.
The App
The app is divided into four main parts: the current theory, the indoor and outdoor cards, the comfort trend, and the daily check-in. Each section below walks through the logic behind that card and what it’s showing. If you’d rather skip the explanation, go straight to the live app below.
Please note that user input data is restricted and only accessible to the primary user.
Current Theory Card:
The app opens with the current theory statement, the AI’s interpretation of what is most likely affecting my eyes right now. The AI (OpenAI) is fed by three sources of data: smart home devices, weather, and user input. That data is stored in a CMS system, and the theory updates whenever I trigger a new analysis.
The Indoor and Outdoor cards
Two category cards, indoor and outdoor. Each shows a circular gauge with a single percentage, how likely that environment is affecting my eyes, backed by a list of individual factors: temperature, humidity, air quality, light level, and screen time for indoor; temperature, humidity, and air quality for outdoor. Each factor has its own bar showing how strongly it correlates with my comfort scores, so I can see which specific condition is driving the overall number.
The User Comfort Trend card
A trend chart plots my comfort ratings across my last 15 logged check-ins, showing whether things are improving or getting worse over time.
User Input
A short questionnaire I fill out once a day: how much my eyes bothered me, what I felt (gritty, burning, watery, blurry, tired), how much screen time I had, and where I spent most of my time, indoors, outdoors, or split, plus an optional note. This is how the AI knows how bad things actually are. The environmental data only tells it what conditions were present; my check-in is the direct signal of how my eyes felt, which is what the AI correlates everything else against.
