Case Study: The Allergy Navigator

I will be fully transparent here: I am building this app for my own selfish reasons. I was at the doctor's recently about a rash, and they had a strong inclination that I have a mast cell degranulation disorder of some kind. Naturally, after I came home, I did hours of research into what the symptoms are, and I realized that this could be the answer I've been searching for my whole life!

The problem? There isn't a treatment or cure for it, other than adjusting your lifestyle. When I looked at the list of potential triggers, to say that I was disheartened and overwhelmed would be an understatement. Sudden temperature changes and fatigue were on the list of potential triggers. I kid you not, my being tired could cause me to become MORE tired because of this mast cell disorder! So I decided to get to work, because no one else will advocate for my health if I don't.

The Intent: Solving the "Diagnostic Delay"

For patients with autoimmune conditions or environmental sensitivities, the path to a diagnosis is often a multi-year "guessing game." The core problem isn't just a lack of information—it's a lack of organized data. These conditions are inherently tricky because the symptoms are so fickle and they often show up in many different ways and in different stages of your life.

The Mission: To build a comprehensive "Health Intelligence" system that tracks the intersection of environmental triggers (food, chemical, beauty, environment) and daily symptom baselines to provide AI-driven pattern recognition for patients and their doctors to use as supporting evidence for a diagnosis.

UX Decisions: Making Complexity Feel Simple

1. Baseline Health

A reaction is rarely an isolated event. A user's "baseline" (sleep, stress, existing flare-up) determines their threshold for a new trigger, and this can drastically change the chances and severity of a flare-up. There are so many small decisions we make every day that could affect this without our consciously even knowing!

• The Decision: Build an easy-to-use, one-tap diary to log new symptoms and habits for the day.

• The Logic: A "Safe" product on a low-stress day might become a "Trigger" on a high-stress day. By weighting the data to include baseline context, the theory is that the AI pattern recognition becomes significantly more accurate at determining when a new trigger occurred versus when it's related to an existing flare-up.

2. Multi-Modal Data Entry (Frictionless Input)

The app needed to handle everything from a bag of chips to a new laundry detergent or a change in seasonal pollen. This is a huge undertaking from a UI standpoint to ensure that it still feels simple to use and not overwhelming. I don't want to have to switch between 10 different modes to log my life, and I'm betting neither does anyone else.

• The Decision: I designed a dual-input system—Instant Scanning with a text, barcode, and photo recognition mode for known products and Manual Ingredient Entry for custom or unlisted items.

• The Logic: By allowing manual entry, I ensured the user is never locked out of their own data, maintaining the integrity of the tracking system even when the database is incomplete.

• The Data Source: To try to seed as much data into the app as possible when beginning this endeavor, I used a few free open-source APIs from the USDA, Open Food Facts, Open FDA, and the EPA (thank you all for making so much readily available data). This populates the app with over 4 million datapoints!

3. AI Pattern Recognition (From Data to Insight)

The most significant feature of this app is the shift from Tracking to Analysis. While the initial v1 logic uses a weighted cross-reference model, I am looking into more robust algorithmic solutions to enhance the AI’s predictive accuracy.

• The Decision: I created a logic flow where the AI cross-references the frequency of specific ingredients against the onset of logged symptoms. It takes into account the symptoms logged in the user's diary for the last 72 hours to determine the chance of an existing flare-up. As you continue to log more data points, in theory, it gets smarter at detecting new triggers.

• The Result: The app doesn't just say "You're sick"; it says "Every time you use a product containing Phenoxyethanol, your baseline drops by 20% within four hours." This is invaluable data for helping you detect patterns in food, environments, or activities that may be increasing the number of days you experience high symptoms.

Advocacy Through Documentation

A major pain point in healthcare is the "Communication Gap" between patient experience and clinical observation. It's hard to sum up how you've been feeling for the last six months into a 30-minute check-up.

• The Feature: I built a PDF Generation Engine that synthesizes weeks of tracking and AI findings into a structured, professional report.

• The UX Intent: This turns the user's "subjective" feelings into "objective" data. It empowers the patient to walk into a doctor’s office with a "Proof of Work" document, directly improving their health literacy and agency. This

The Knowledge Ecosystem

To close the loop on health literacy, I integrated a research-backed Wiki.

• The Design Goal: To provide a centralized, easy-to-navigate repository of the latest findings on autoimmune and environmental health, including any publicly available scientific papers, new studies, and any trials or breakthroughs in the research.

• The UX Logic: Education is the first step in symptom reduction. By placing the Wiki alongside the tracker, the app moves the user from "Crisis Management" to "Proactive Prevention." I believe the more educated you are, the better equipped you are towards a recovery or at least a better quality of life.

The Retrospective: Designing for the Patient

What started as a "selfish" quest for my own health answers has evolved into a blueprint for a new kind of healthcare advocacy. This project taught me that Information Architecture isn't just about organizing a website; it’s about organizing a life.

By seeding the app with over 4 million data points from the USDA, FDA, and EPA, I’ve moved the project from a simple diary to a robust Environmental Intelligence Engine. The goal isn't just to track triggers, it's to reduce the cognitive load of existence for people whose bodies have become high-friction environments.

What’s Next?

The logic for the AI pattern recognition is the next frontier. While the current model works, it's still pretty basic. The future of this system lies in deeper Machine Learning that can account for the compounding effects of "Baseline" stressors over months, not just days.

There are many bugs to fix and features to polish up, as this is just the MVP product for my own personal use. But as I began developing it, I felt that this app could be extremely useful in the quality of healthcare and advocacy for the majority.

As time allows, and if interest grows for this app, I will continue to improve upon my initial design to make the UI look and feel better to use and easier to understand.

Ultimately, this app is a testament to my philosophy of Operational Minimalism: building the most sophisticated systems possible so that the human using them can finally do less work just to stay afloat.

If you're interested in being one of the initial testers of this app, you can send me an email with the subject line Allergy Navigator to info@hillaryedenmcmullen.com,

and I will add you to my testing group. Currently, it is only available for Android phones (sorry, Apple users!)