Inspiration
Grocery store runs are a hassle, especially when you're shopping in a new environment. ShoppingSpree streamlines the process with a simple UI/UX that gives you the shortest route to pick up your groceries.
What it does
ShoppingSpree allows store-owners to create a floor plan of their store with labels of where different aisles/sections are (Meat, Rice, Pasta, etc.). We allow users to enter a grocery list, use an LLM to classify the items in the grocery list by aisle, and use A* shortest path algorithm utilizing the floorplan from the store owner.
How we built it
The frontend is pure HTML/CSS and JavaScript interactions. We call a free LLM API from HuggingFace. We are using Firebase for database management on the backend.
Challenges we ran into
Limited experience with databases and backend development.
Accomplishments that we're proud of
Smooth API interaction, nice-looking frontend, well-designed architecture.
What we learned
We greatly enhanced our skills in web-development through the creation of this project.
What's next for Shopping Spree
Integrate a stronger LLM (GPT-4) where error rate in classification will be near zero. Allow users to import grocery lists from popular retailers like Wal-Mart or H-E-B. Clean up UI/UX. Improve authentication and database security.
Log in or sign up for Devpost to join the conversation.