doc4<fn (Winner of Best use of AI)

Inspiration✨

When Jeremy was walking to our house to work on the hackathon, he accidentally slipped on ice. Fortunately, he did not face a major injury, however, he had feelings of discomfort. This sparked the idea that what if Jeremy could simply visit an online doctor for this physical injury, without going through the trouble of leaving the house? Additionally, a heavy snowstorm was reported to occur in a few hours, in which everyone was recommended to stay at home.

What it does 🍎

The website allows users to first chat with either a doctor or an AI to receive a diagnosis. Using the chat option, the website will ask users to describe their injury so they can accurately diagnose the user with rehabilitation exercises. After, the user can use the track option to track their progress and range of motion whilst performing the exercises. This allows the user to ensure that they are doing the exercises correctly in order to heal accurately.

How we built it 🔧

The website was built using various software such as javascript and express.js. The front end was created in HTML5 and CSS3, while all the icons were self-drawn in Adobe Illustrator. The chat option was created with socket.io and openAI. The track option was created with TensorFlow, ml5, and posenet.

Challenges we ran into 🚧

The challenge faced when creating the functionality of the website was getting socket.io set up to process the chats. It also took a while to come up with our own unit of measure that determines one's range of motion using joints as well as displays live data on a graph.

Accomplishments that we're proud of 🏆

We are proud of the chat option since we were able to connect the openAI database to accurately reply to the user. We were able to overcome and correct the many errors that occurred during the process. We were also incredibly proud that we incorporated TensorFlow.

What we learned 🧠

We learned the most from the wild dynamic of the hackathon and how everything suddenly became virtual. This taught us to be able to adapt quickly to the situation at hand in order to minimize disruption in workflow. We decided to all work in one house rather than online to simulate the in-person hackathon experience.

What's next for doc4<fn ⏭️

This website has a lot of future potential where it can essentially expand towards pharmaceuticals and more medical services, e.g mental. We can also partner with medical and technological companies to provide medications and flourish in the industry.

Built With

• adobe-illustrator
• capcut
• css3
• express.js
• html5
• javascript
• ml5
• openai
• posenet
• socket.io
• tensorflow