DermaDial | Devpost

Inspiration

Skin conditions affect millions of people globally, yet many go undiagnosed due to lack of access to dermatologists or fear of seeking help. We were inspired by the idea of using technology like a time machine—to fast-forward the long, stressful wait for a skin diagnosis. Instead of days or weeks spent booking appointments, waiting for results, or researching symptoms, we wanted to give users immediate answers and guidance. With DermaDial, we aimed to bring the future of dermatology to the present—instantly, and in the palm of your hand.

What it does

Users simply take a photo of their skin condition, and our AI model instantly analyzes the image to identify the likely disease. It also includes an AI chat assistant that provides quick, reliable information about the diagnosis—no waiting rooms, no delays. It’s fast, accurate, and ready whenever you need it.

How we built it

We used Teachable Machine to train our AI model on a set of skin disease images. Then, we integrated that model into the app with a clean interface designed for speed and simplicity. Firebase was used for secure user authentication. We also created an AI-powered chatbot to act as a virtual dermatologist, answering user questions in real time. We used the Gemini API Key in our AI Chatbot. Finally, we built a website to promote DermaDial and explain the features.

Challenges we ran into

There were many challenges in trying to build DermaDial. First, we could not hot reload our app due to the WebViewWidgets stacking on each other, which would cause an error. To work around this, we had to not hot reload the app. We also ran into some errors with Teachable Machine and had to decrease the amount of data for each type of skin disease. For our AI, we also needed many images and websites like kaggle didn't have all the data we needed. That is why we used data augmentation to work around this and get all the data we need. We used a python script to rotate the image 30 different ways.

Accomplishments that we're proud of

We were very proud of how we were actually able to finish this project and were happy we made an app that includes many AI features. We are also proud that we made our own website.

What we learned

We learned how to integrate Teachable Machine with Flutter and learned how to create an AI chat. We also learned to make a website using Flutter.

What's next for DermaDial

Expand the AI model with more skin conditions and a larger dataset for improved accuracy.
Partner with dermatologists to verify and increase reliability.
Launch a mobile app to the App Store, bringing our App to a variety of users.
Add features like progress tracking over time, so users can monitor how their skin changes.
We also want to add more to our website, such as google maps integration so users can find a Dermatologist.
Support an offline mode for users in remote areas.