TerpPlate

Inspiration

As students who who eat on campus frequently, we have notice how easy it is to eat mindlessly when dining hall food isn't displayed with nutrition labels. We wanted to create a tool that empowers students to make informed food choices that align with their health and dietary goals.

What it does

TerpPlate is designed to generate healthy meal ideas based on available dining hall options, making it easier than ever for busy college students to select balanced, nutritious meals. Users can specify their calorie, sugar, sodium, and protein preferences, and TerpPlate will generate a meal suggestion that meets those goals. With daily data on all available dining hall foods, each meal is curated specifically to what’s being served that day, so you can always plan your meals accordingly.

How we built it

To build TerpPlate, we developed a Python web scraper to gather menu and nutritional data from UMD’s Dining Services website, organizing it into a structured format for our AI model. Using the Langchain4j framework and OpenAI’s GPT4oMini model, we implemented AI components that generate balanced meal suggestions based on user preferences and a Nutritional Chatbot for dietary guidance. The front-end, crafted with HTML and CSS, delivers a clean, responsive user experience, while the backend, built with Spring Boot, ensures robust handling of user requests and data processing. Maven was used for managing dependencies and project configuration, creating a seamless and efficient development environment.

Challenges we ran into

The largest challenge we ran into was trying to integrate the many different programs and frameworks without simple tools. For all of us, this was our first time working on a project of this size and complexity and so it was a challenge to bring everything together.

Accomplishments that we're proud of

We're proud of our ability to make the UI similar to the style that UMD uses for its dining services. We are also proud of the way we were able to make use of a new frameworks we had not used before and integrate the individual parts together.

What we learned

We learned about the different aspects of working collaboratively, both on the technically and interpersonally. Given the multidisciplinary nature of our project, we were able to utilize git version control along with our teamwork skills to make a cohesive project.

What's next for TerpPlate

In the future, we would like to improve our application by adding more features to further personalize the meals for the user by filtering for dietary restrictions. We would also like to add an option for the user to log their meals in the app..... Lastly, we look forward to enhancing our frontend by adding more features to improve user navigation and improving the look of the application.