GrocerScope

User Input: zipcode and desired item
Results for desired item (near the user's zipcode) sorted by price
The built-in Gemini chatbot answering a user's question

Introduction

Many of us have faced the agonizing choice of wasting your day in search of cheap groceries, or giving up and settling for one of the many price-gouged options.

Now, gone are the days where you have to sacrifice either your money or your time. With GrocerScope, you can optimize your shopping straight from your couch.

What it does

All you have to do is input your zip code and an item you're searching for. It'll do the rest to give you the cheapest available, in a location convenient to you. Not sure what you want? Just like browsing an aisle, you can search in broad categories like "dairy" to experience all the possibilities your area has to offer.

How we built it

GrocerScope is mostly built in Python, with some HTML/CSS/JS for the user interface. The frontend takes three user inputs and presents a Gemini-based chatbot to help guide selections. The backend uses a web scraper to find location and pricing data from store websites. We used Django for both in order to maintain continuity and ease of integration.

Challenges we ran into

Early on, we had a lot of problems with web scraping. Many grocery store catalogs were difficult to parse, and our implementation was buggy. We were eventually able to overcome this by carefully analyzing the scraper's inputs and outputs, and combining those insights with AI tools to discover sources of error.

Our original idea was actually to do a cost comparison for delivery apps. However, due to high price variability in these apps, it was much more accessible to use publicly-available grocery store data. Since people typically go to grocery stores more often than they order delivery, we decided this new approach maintains our mission of helping people save money.

Accomplishments that we're proud of

Firstly, overcame technical challenges such as function match errors and poor web servers from the stores.

Moreover, we made something that is useful and helps the community. With the help of our app, everyday people can save money in a way they may not have envisioned before.

What we learned

Before BostonHacks, none of us had ever met each other or participated in a hackathon. After the end of these 24 hours, we had learned a lot about web development, full-stack integration, and effectively collaborating with AI to streamline creation. All of us have come out of this experience more knowledgeable about the tools used by real software engineers, and more confident in our ability to handle a changing world.

What's next for GrocerScope

GrocerScope is currently most useful in the Northeast due to the franchises it includes. The next step is to develop partnerships with more grocery stores across the nation. Additionally, communicating with a diverse array of companies can help us achieve our original goal of comparing food delivery apps, perhaps worldwide.