Caduceus

• 

  our logo

• 

  homepage w/ chat

Inspiration

In Southeast Asia and many other developing countries, one of the biggest problems that an average citizen may face is getting treated for their injuries, diseases, and more. Healthcare is either too expensive, too inaccessible, or too inefficient. As a group composed of first-generation, low-income students with families still in these home countries, we are often told stories of the many health problems that are families unfortunately have to face due to the healthcare system in place. As a result, we wanted to build an app that focuses heavily on accessibility and utility for users in developing countries, inspired by the heavy fight that our families go through.

What it does

We have built a chatbot powered by Gemini that leverages its text-to-text and computer vision capabilities to help give users more information about their injuries, as well as how to treat and further prevent it. We have also used multiple APIs to ensure that people can communicate with the chatbot in different languages, opening arms to users who do not know how to speak English.

How we built it

Frontend Stack:

• Flutter
• Dart

Backend Stack:

• Gemini API
• Google Cloud Translate API
• Langdetect API

Challenges we ran into

Challenges we ran into when creating Caduceus was finding the information necessary to determine the problems that are common within Southeast Asian countries and what steps we could do to help give the people some type of aid. Providing aid other than seeking healthcare professional is difficult, but with the addition of a camera feature to provide a more precise answer, both seeing and describing would make it likely for the chatbot to provide more personal and effective treatments. Another issue arose when finding out that the majority of the database used Android, leading us to use Flutter for the first time in order to cater to that demographic and not just the Apple users.

Accomplishments that we're proud of

While developing our app, some accomplishments we were most proud of was being able to successfully complete our backend using Gemini API and Google Translate API. This task was huge to our purpose because leveraging these new tools can have a huge impact on our users. Developing our front-end was also a massive hurdle that we were proud of overcoming due to its seemingly steep learning curve, however, it was very rewarding.. Being able to put the project together in one uniform concept was our biggest achievement because communication is also key to building a successful project and our goal is to communicate with our users.

What we learned

Setting up the environment and emulating our app live onto our devices was very rewarding to experience and we learned a lot from our time during this hackathon. Learning to use the flutter SDK was very interesting as its compatibility on both IOS and Android made it a useful tool to have in our back pockets. Along with learning how to implement the challenging Gemini API and its tricky computer vision, we now have a powerful ally at our hands when developing AI-powered projects.

What's next for Caduceus

Caduceus plans to retrieve more data regarding each country to provide better responses to each country’s common issues or to understand the user’s personal data and connect the problems they’re listing in order to possibly identify an underlying problem that may not be seen at first glance. Caduceus will continue to evolve, as improving the application to have more features such as a camera and the chatbot providing responses is an ideal product for our users.

Built With

• dart
• flutter
• geminiapi
• google-cloud-translate
• langdetect