The app uses Google Gemini's visual API to scan the users face based on their age, sex, and any allergies. It analyzes the users skin type to determine any visual flaws, and from there recommends a skincare regimen based on a users inputted budget.
This app was purely built as a means of testing the limits of Google's AI studio platform. The platform was built with TypeScript and Gemini API, with minimal changes to code needed as majority of the work was done through fine prompting.
With designing code in pure english, the primary issue came with changing specific aspects to our application. In most cases, we'd have to edit the code directly through the program files.
The speed and efficiency at which we brainstormed, prompted, deployed, and manipulated the application is definitely the biggest accomplishment from this development process.
There was a lot of knowledge gained from this process, the most significant being what you can build with AI, finding its limits, and when you need a human to take over the wheel again.
Ideally the next implementation is focusing on validating the products that the AI engine recommends, through validation of links or other means. There's also hopes of implementing a backend structure, where users would be able to save their results and view improvements or otherwise through time.
Log in or sign up for Devpost to join the conversation.