[ sorry for the horrible video and audio quality ]

Inspiration ✨🌟

My Inspiration was a real problem by hospitals these days:

  • There is no room allocation system properly, which leads to delays.
  • Hospitals think they have the proper rooms sometimes, but when an emergency patient enters, and there is no room available, that can lead to a life being lost.
  • There is no proper patient record.
  • Sometimes, the 911 or 112 caller doesn't know what the problem user has, or it's an accidental call or what, but if the person goes to this website, and click this, the location will be automatically shared with them, in a voice message by Twilio saying that there's an health emergency, my initial idea was to use google maps API to get the phone number of the nearest hospital to call for ambulance.

The project solves these problems ^^

What it does 🏗️👷

It does the following:

  • It helps to manage patient records.
  • It helps to call for emergency, talking as a real human using Twilio automation api.
  • It helps to automatically allocate rules, automate alerts/notifications based on the sent ones manually, and the ones sent using AI AND MACHINE LEARNING MODEL THAT IS USED TO DETECT if the person is on a stretcher or a wheelchair, that can also help in already telling if a room is available or not, to NOT , NOT AT ALL put any delays, and that can potentially save a life.
  • It also allows the people to put their medical info, that can be downloaded as csv, and used to train a model for prediction/research purposes.
  • It uses Twilio API, and firebase.

How I built it 🏗️🔥

This is how I built it:

  • I used NextJS, Twilio API, Firebase Auth, and Firestore to build this website!
  • I used Github for project management, render.com to host the api.
  • I used Tailwind css to style stuff!
  • I used Google's Teachable Machine Learning thingy to create a model.
  • I also used the motivation from office hours! 🔥

MODEL LINK: https://teachablemachine.withgoogle.com/models/IMbZYBBnc/

accuracy is less because of less dataset.

Challenges we ran into 🚫🤚

There were like a billion challenges that I faced:

  • time problem, I was only restricted to use laptop for a few hours, but I used the laptop unethically to fulfill this challenge!
  • homework from the classes, 7-8 hrs in classes, 3 hours of nap, 8 hours of sleep, 3 hrs of homework, and the rest for this project!
  • Unable to find a team :(

Accomplishments that we're proud of 🏆✨

I'm so happy that I'm able to finish this project, and I learnt about Twilio API, and teachable machine learning for the model!

What we learned 📚🏫

I learned:

  • NEVER GIVE UP!
  • Everything will work out if you try!
  • Learnt about Twilio API, learnt about teachable machine learning, learnt about firebase firestore in NextJS, learnt about render.com.

What's next for MedAlert 📊📈

My Plans for this website:

  • Enhancing dataset/model, and adding more categories.
  • Better room allocation.
  • More classes.
  • State management, to not make it laggy like it is right now.

Thanks everyone who read this and hosted the hackathon, judges, MediHacks team, viewers, and everyone 🏆💖

Built With

Share this project:

Updates