Allenu

Inspiration

Food allergies are an oft-overlooked but extremely common chronic hidden disability. Studies have shown they may be linked to higher rates (between 1.6 and 2.3 times more likely) of depression, stress, and anxiety, leading to impaired quality of life. One in thirteen children has a food allergy, and 40% of children with one food allergy also have more, leading to increased stress for guardians as well. It's an affliction with possibly devastating acute effects, but chronic stress leads to definite sinister long-term effects.

The simple act of eating at a restaurant can be fraught with danger. Ordering food requires a person to remember each allergy and its severity, while the waiter must remember the possible allergens in each dish. That's a human and fallible process, but with Covid-19, restaurants are moving to online menus tied to QR codes, adding another hurdle to the process.

What it does

This is where Allenu comes in. Allenu is an allergy-aware online menu. Users can import their allergy data from external health providers and view menus from participating restaurants. Each potentially dangerous menu item is highlighted with its relative potential severity, saving the error-prone and arduous processes outlined above.

How we built it

We built it by leveraging InterSystem's IRIS data store and API. IRIS let us quickly deploy an instance of a health provider on AWS, and we then populated it with seed users with their own sets of allergies. IRIS acts as both an identity and a health store. When a user first visits Allenu, they log in by providing their IRIS ID, and then we make a GET request to IRIS to load their name and allergies. When a user visits a menu, we load the menu items from our own database, load their allergies from IRIS, cross-reference the two, and display a menu with appropriately dangerous dishes flagged. Users can also add and delete allergies because those requests are passed through our system to IRIS.

Our IRIS deployment is running on AWS. Our web app is built using Flask and uses PostgreSQL as a database, all hosted on Heroku. We also use Heroku for secret management.

Challenges we ran into

Becoming familiar with the IRIS API took some time, but we had some engineers at the InterSystems sponsor booth walk us through the deployment process and API reference material.

Accomplishments that we're proud of

We're proud of using IRIS because it gives us the data in FHIR format. FHIR is an industry-standard format for health data, meaning our app could easily extend to using any external health provider. We're also proud of incorporating functionality for the user to add and delete allergies by passing them through to IRIS, giving the user stewardship of their own health.

What we learned

The big takeaway for us was the sheer scale of the mental health impacts of allergies, along with their silent ubiquity. On the technical side, we learned how to integrate data from both remote services and local data stores to serve complex user queries.

What's next for Allenu

We haven't won yet! We're looking into possibly automating data ingestion with web scraping so it's even easier for restaurants to add their own menus. Integrating with other health providers would require a user data authorization flow which we'd also like to build.