AllerSense

Search Database
Update Profile
Past Scans
Homepage
Scan Items

Inspiration

For those with more common allergies it is very easy to find allergens on food labels, but for those with less common allergies, this information may not be so easily apparent. These allergies may be hidden in a sea of other ingredients, or not even listed because it is a derivative ingredient. We created AllerSense to help those with less common allergies easily identify foods they are allergic to.

What it does

AllerSense utilizes computer vision to parse ingredients lists and alert the user if the item they scanned contains something they are allergic to. AllerSense also uses databasing technologies to allow users to search for certain foods and see if they are allergic to them without uploading images.

How we built it

We built AllerSense by using Next.js with Typescript and Tailwind CSS for a powerful and dynamic frontend interface. For the backend we used flask to handle all our computer vision related needs, uploading images, and scanning text. For the database we leveraged the power of Google Firebase to have a flexible and adaptive no-sql database for all our data storage needs. Finally, we used Propel Auth for authentication to ensure a secure and reliable user login/sign-up.

Challenges we ran into

Initially, we wanted to build a search engine to quickly query the user inputted foods from the database to ensure scalability as the user base grows. Despite our ambitions, this was simply too difficult to complete in one day, so instead we opted to use firebase so that we could get the base functionality covered. Another issue we faced was uploading, and cropping the image in the frontend and sending it to the backend to be processed. This was very time consuming and difficult to figure out.

Accomplishments that we're proud of

We are proud that we managed to handle all the database logic, allowing for separate users with their own individual and personalized data. Another thing we are proud of is managing to upload the images to the backend for processing, since this was a very technically challenging task.

What we learned

We learned to plan out project infrastructure more carefully before we start coding to ensure a smooth developing experience, without having to pivot and change directions too much.

What's next for AllerSense

We want to expand from allergies to other dietary restrictions, such as vegetarian, vegan, kosher, and halal. We also would want to be able to feed the ingredients list into a model to find out more information about derivative ingredients.