Inspiration
Growing up, my mom was a nutritionist and I became very well-versed in food sciences. Coming to college, I was surprised that most people have very limited exposure to nutrition and just get their information from short-form content, which often times is contradictory and not personalized. I wanted to create something accessible for everyone to use that considers your nutritional needs and your goals to provide advice on what can you help you achieve what you want.
What it does
The app itself is simple-You are recommended to enter your height, weight, sex, lifestyle, and what lens you want to view with. Then you search up a food/dish and the app provides recommendations and scores of the food you searched from different criterion, and also provides alternatives.
How we built it
We used Swift for the UI with an Node/Express backend that calls the Claude API. When the user inputs information about themselves to customize their search, it formats it into a prompt that is used to call the Claude API
Challenges we ran into
We were both very new to iOS development, and working with XCode and Swift was completely new for both of us.
Accomplishments that we're proud of
Working an MVP that actually solves a problem that we personally face, that can extend to everyone. Health and nutrition is something I'm very passionate about.
What we learned
The main aspect we learned was that simplicity and being straighforward is more effective with conveying our purpose. We were initially stuck on wanting to make something technically complicated, and honestly users don't really care about your tech stack, as long as you app works. We kept our UI and backend simple, and focused on making the app concept easy to understand and the interface easy to use.
On the technical side, we learned how to work with different editors to use Swift, Express Backend, while using the Claude API. Specifically, we spent a lot of time debugging through git logs and resolving issues with bad commits. We gained a much stronger understanding of collaboration through source control.
What's next for Nutrition Helper
Ideally we would want to have an actual database that stores common searches and includes more in-depth analysis from health studies. Another further extension would be being able to directly search up products and gather the same personalized data for that specific product.
Log in or sign up for Devpost to join the conversation.