Wolfieats

Homepage of application where you can choose what meal plan you have
Prompt page where users can ask food related questions
Prompting bot for vegetarian options on campus

Inspiration

We like to eat. Who doesn't? So, we made it easy to find food that satisfies our cravings.

What it does

Wolfieats essentially serves as a chatbot that, based on user input, will help indecisive or curious Seawolves subdue their growling (or soon to be growling) stomaches. Wolfieats considers your meal plan as a Seawolf and from there on out optimizes the best course of action for your next meal based on your needs and cravings.

How we built it

We used Next.js to build the frontend of the project, while using a python script to retrieve data on the food choices available on Stony Brook University's campus. We then trained an AI model on said data in order for it to make precise and accurate responses.

Challenges we ran into

Obtaining the actual data of all the food available on campus was an apparent challenge we instantly ran into, as it took way more manual labor than expected to fetch and properly format the information in order for the data to actually be usable.

Accomplishments that we're proud of

We're proud that we were actually able to put a mere random idea into a physical application that can be used.

What we learned

We learned and expanded our knowledge on skills such as Next.js, Tailwind, and how to direct an AI model in the right direction with prompting and tailored data. We also learned that obtaining and preparing data is a lot more tedious than expected and is not a fun thing to do.

What's next for SBU Food Recommender

Future steps include fine-tuning the model and train it more for even greater accuracy, expanding the food scope to off-campus but nearby eats for greater diversity, and speeding up the response time of the bot.