Goose Grocer

Home Page + Flyer Parser
Grocery List Maker
Product Comparison
Meal Planner + Random Gen
Meal Breakdown
Bulking Preparation
Database
Weekly Foods
Daily Meal Planner
Shopping List Breakdown + List Downloaded

Inspiration

Optimizing grocery shopping can appear daunting to students used to living on campus after first year—at least it does to us. Prices change, flyers are abundant and messy, and there are a surplus of options to an overwhelming degree. We noticed that most people either shop at one store out of habit or manually jump between flyers without an optimized or easy way to compare.

We wanted to build something that turned these time-consuming activities into a simple process: type what you need, and immediately see where it’s cheapest. The idea was inspired by the needs of the average university student: affordable food, quick to use, and health-conscious.

That’s how Goose Grocer was born: a slightly chaotic, very practical grocery comparison tool optimizing your grocery shopping and meal planning.

What it does

Goose Grocer helps users find the cheapest groceries to buy at major grocery stores and chains.

Users can:

Enter a grocery list or search individual items
Compare prices across multiple stores in one place
See flyer deals alongside regular prices
Build and adjust a smart grocery list
Plan their meals across the week
Generate recipes and estimated meal costs

The core idea is simple: an all-in-one app centered around students feeding themselves.

How we built it

We built Goose Grocer as a lightweight web app using:

Python for data processing and business logic
Streamlit for the frontend UI
Pandas for cleaning, merging, and comparing price data while offering a clean display
SQLite/MySQL-style database to store products, flyer deals, and saved user data
Gemini AI for recipe creation and optimization tasks

The app structure separates concerns clearly:

A database layer for products and parsed flyer data
A comparison engine that normalizes prices and handles edge cases (missing data, multiple matches, unit differences)
A Streamlit interface that keeps everything interactive and demo-friendly

We focused on shipping a working end-to-end flow first, then tightening logic and UX once the foundation was solid.

Challenges we ran into

Data: Acquiring reliable data was difficult. We started with a seeded dataset generated by ClaudeAI, then manually scraped sites like Walmart to ensure accuracy.
Merge errors: With four relatively new programmers collaborating, merge conflicts were common.
Streamlit limitations: Simple UI tweaks sometimes required long workarounds. Streamlit was quick to start but tricky for complex UI.
Debugging: Forced us to truly understand each other’s code and assumptions.

Accomplishments that we're proud of

Building a fully functional grocery comparison engine from scratch
Creating a method to account for flyers and receipt scans
Designing an app that’s intuitive and demo-friendly
Most importantly, solving a real problem we personally face

What we learned

This project taught us more than just tools and software:

Clean data matters more than fancy features
Small UX decisions can drastically change usability
Writing modular code early saves massive debugging time later
Shipping something imperfect beats planning something imaginary
Greater confidence working with APIs and implementing AI in projects

What's next for Goose Grocer

Next steps include:

User accounts and saved grocery lists via Auth0 (attempted but cut due to time)
Recipe-based shopping with automatic ingredient expansion
Location-based store recommendations and filtering
Receipt comparison (scan + Gemini Vision) to suggest better stores next time

Long-term, Goose Grocer could evolve into a full student-focused grocery assistant—but only if it keeps doing one thing extremely well: making groceries cheaper with minimal effort.