HealthHack

All around the world, people are faced with many challenges that prevent accessibility to healthcare. In our project, we create a virtual assistant that helps maintain better health using LLM.

Comment

Inspiration

Millions of people struggle to access the basic healthcare they need, especially in rural regions and for low-income individuals. Our team has tackled these issues that many individuals encounter and has developed a creative solution that will assist them in taking care of their health and gaining access to essential services.

What it does

It acts as your virtual friend, which helps you with self-diagnosis and gives you home remedies before consulting a doctor. It also helps you provide the available services, like CalFresh, to get healthy on budget.

How we built it

This project is divided into two phases: Front-end: It is the webpage that has the chatbot interface and is built using CSS/HTML. Backend: We have the LLM (BERT) model, which is trained on more than 300 different unique diseases, predicts the type of disease, and gives the required remedies. In order to make it easier and more user-friendly, we have used a Google Translator that converts the English output into Spanish.

Challenges we ran into

Our team is a mix of front-end developers and data scientists. We were able to understand, analyze, and interpret meaningful insights from the data. Build models to make the predictions. We also build our front-end UI. Our team faced challenges in integrating the models into the front-end UI.

The data collected was a real-time data extraction method, which was time-consuming.

Accomplishments that we're proud of

  1. We were able to predict the type of disease and provide the remedies using the LLM Model.
  2. We were able to project the results in multiple languages taking users language factor into consideration.
  3. We finally built a prototype of our Chatbot with sample data inputs and output.
  4. We were able to create the login pages of the website using HTML/CSS.

What we learned

  1. This project gave us an opportunity to come out of the box and think something different.
  2. Effectively work as team and collaborate with each other.
  3. We were able to breakdown the large problem statement into multiple tasks and work on the approaches. ## What's next for HealthHack
  4. We plan to integrate the chatbot with the backend LLM model.
  5. We plan to get the google maps api and embed it to our model so that it can actually recommend users the nearest hospitals available.
  6. We plan to include the resources like Calfresh, California Association of Food Banks, School Meals etc into our chatbot which will give users knowledge on the vital services they can use to stay healthy.
Share this project:

Updates