Inspiration
How many times have you forgotten to take your medication and damned yourself for it? It has happened to us all, with different consequences. Indeed, missing out on a single pill can, for some of us, throw an entire treatment process out the window. Being able to keep track of our prescriptions is key in healthcare, this is why we decided to create PillsOnTime.
What it does
PillsOnTime allows you to load your prescription information, including the daily dosage and refills, as well as reminders in your local phone calendar, simply by taking a quick photo or uploading one from the library. The app takes care of the rest!
How we built it
We built the app with react native and expo using firebase for authentication. We used the built in expo module to access the devices camera and store the image locally. We then used the Google Cloud Vision API to extract the text from the photo. We used this data to then create a (semi-accurate) algorithm which can identify key information about the prescription/medication to be added to your calendar. Finally the event was added to the phones calendar with the built in expo module.
Challenges we ran into
As our team has a diverse array of experiences, the same can be said about the challenges that each encountered. Some had to get accustomed to new platforms in order to design an application in less than a day, while figuring out how to build an algorithm that will efficiently analyze data from prescription labels. None of us had worked with machine learning before and it took a while for us to process the incredibly large amount of data that the API gives back to you. Also working with the permissions of writing to someones calendar was also time consuming.
Accomplishments that we're proud of
Just going into this challenge, we faced a lot of problems that we managed to overcome, whether it was getting used to unfamiliar platforms or figuring out the design of our app. We ended up with a result rather satisfying given the time constraints & we learned quite a lot.
What we learned
None of us had worked with ML before but we all realized that it isn't as hard as we thought!! We will definitely be exploring more similar API's that google has to offer.
What's next for PillsOnTime
We would like to refine the algorithm to create calendar events with more accuracy

Log in or sign up for Devpost to join the conversation.