Frugal Express

Inspiration

As young adults, it can be challenging learning to cook to sustain yourself while attending school. An added challenge is buying groceries on a student budget. After a long day of attending classes and studying, one of the most hated questions is: "What am I going to eat tonight?". We wanted to help reduce some of our frustration by "cooking up" a project that provides a quick and easy way to shop savvy and get inspired in the kitchen.

What it does

Our project allows people to ask questions and access recipes that match their query. You can ask for a recipe with specific ingredients, number of ingredients, under a certain price, and more. Try it out!

How we built it

1. Use 1M+ Recipes dataset
2. Web Scrape Grocery Store Data with UIPath
3. Build a Grocery Item Dataset
4. Process both datasets using NLP, stop words, data science techniques, etc. to find connections
5. Create a scoring/processing algorithm to correctly process the recipes data as well as linking both datasets into a single combined recipe dataset
6. Upload all data into MongoDB Atlas (Very big)
7. Create a React Native Mobile app to host
8. Create a Flask Server
9. Use GCP Dialogflow and Speech-to-text to communicate with the Flask server

Challenges we ran into

1M+ recipes is very challenging to process all at once (the file is 2Gb!) Grocery stores have really bad UI which make scraping content very challenging (page crashes halfway)

Accomplishments that we're proud of

Overcoming many challenges to create two different datasets and connect them together

What we learned

Dialogflow Data Science Techniques Natural Language Processing The challenges of Big Data (especially in a short time limit)

What's next for Frugal Express

Making use of our domain: thegrocerystoreishacked.online

Built With

• atlas
• dialogflow
• mongodb
• natural-language-processing
• react-native
• uipath