Inspiration
Staying healthy is a goal that everyone strives to achieve, but sometimes life just gets in the way. For the average person, slacking off for a few days is not a big deal and is something that is often done. However, for the more vulnerable population, staying on top their health is crucial and can often feel like a burden.
In 2017, a study looked into people's ability to adhere to their medication. This study found that 80% of the time individuals delayed taking their medication, while 45% of the time they completely forgot!
We set out to create a personalized medication management app that would be both accessible and easy-to-use app. We wanted to specifically focus on the more vulnerable population (those with disabilities: Alzheimer's, severe arthritis, etc) of senior citizens to easily keep on top of their health.
Another common issue we found from anecdotal experience, is that senior citizens often dislike using technology and find it to be overwhelming. We wanted to bridge this gap and make the user experience for seniors as seamless as possible and take away the burden of unfamiliarity.
What it does
PrescribePal is a smart health app with an emphasis on accessibility. Its main features are:
- Creating prescription order requests using OCR and voice interactions
- Giving friendly reminders to users to take their medication
- Easily connecting patients with professionals at a moment's notice
PrescribePal allows the user to opt-in to receiving timely notifications throughout the day or week. These notifications remind users to take their specified medication and amount.
Our app will also automate the process of refilling medication and supplements. Users can upload pictures of their medication to our app, and trust that PrescribePal will be able to automatically refill that medication upon request. Upon making a request, PrescribePal will also send out a text message to a caregiver or doctor for the appropriate approval.
Finally, users will be able to get in touch with their medical professionals at the click of a button. Our app will feature an online meeting platform where users can meet virtually with their doctors.
How we built it
We built PrescribePal using the MERN stack as well as the following:
Socket.io and peer.js for video calling
Tesseract.js to extract text from images taken by patients
Cohere.ai to tokenize large chunks of text found on prescription labels
Puppeteer to find relevant product data
Alan.ai to create an easy to use voice assistant for patients
AWS to host and fetch our images
Challenges we ran into
Trying to integrate our project with various different API's and libraries proved to be a difficult challenge. Working to integrate the entire end-to-end flow of our app starting from an image and transforming that into request for medication as well as connecting with Amazon. Our team barely had any prior experience working with these technologies in small applications, let alone creating a full scale project. Additionally, learning to work with so many requests and passing them between different libraries was extremely time consuming. After a significantly underestimated amount of effort, our team was finally able to integrate our Tesseract.js image conversion, Cohere AI text tokenization, Puppeteer.js web-scraper, and Twilio SMS notifications all into the desired degree of completeness.
What we learned
Having spent so much time working with API's and formatting requests, our team can confidently say we are much more experienced. Understanding how to properly parse through hundreds of lines of JSON objects was a big lesson. Team members each got their own opportunity to refine and improve upon previously lacking skills. Further exposure to front-end development was a learning experience for one, and for the other it was greater understanding of push notifications from a server-client aspect.
What's next for PrescribePal
To further improve the functionality and accessibility of PrescribePal:
- We'd like to improve the user interface to make it even more accessible for users
- Create a more personalized virtual video calling room for doctors with features like note taking, text chat and white boarding
- Reduce latency on notifications being received by patients for a more seamless experience
Built With
- alan
- amazon-web-services
- cohere
- domain.com
- mlh
- mongodb
- node.js
- ocr
- react
- twilio
Log in or sign up for Devpost to join the conversation.