Description as a Tweet:

Our aim is to offer an alternative to expensive and inconvenient Smart Fridges. Our product is a cheap kit which can turn any fridge into a smart fridge, and with features most conventional smart fridges don't have such as automated video classification of inventory.


I went to the Amazon Go store this summer and it made me realize how silly entering in things by hand is.

What it does:

Have you ever forgotten anything at the grocery store? Or better yet, bought items that you already have?
Writing a shopping list can be difficult, and checking to see everything you need is a long tedious process, that still has room for error.
New smart fridges are too much money $$$$
You are paying for too much unnecessary functionality!
There needs to me some in-between

We have a solution. Our simple hardware and service allow you to manage your fridge inventory without ever entering food into a computer. In the age where the govt tracks your face, why can't you keep track of your food?

How we built it:

We built a classifier on a Raspberry Pi (which did not finish installing in time) and our service is architect ed in the AWS cloud using Route 53, Lambda, S3, and DynamoDB.

Technologies we used:

  • Javascript
  • Node.js
  • React
  • Python
  • Raspberry Pi
  • AI/Machine Learning
  • Other Hardware

Challenges we ran into:

The Raspberry Pi being so difficult to install python libraries on.

Accomplishments we're proud of:

Architecting a hugely scalable service in the cloud

What we've learned:

Not to trust Raspberry Pi for quick projects

What's next:

Scale it up and increase recipe api effectiveness and give smarter recommendations

Built with:

Our laptops and AWS

Prizes we're going for:

  • Best Hardware Hack
  • Best Web App
  • Best AWS Hack
  • Best Domain Registered with

Prizes Won

Best Use of Amazon Web Services

Team Members

Cole Smith
Kevin Gill
Owen Carpenter

Table Number

Table 71

View on Github