PantryLens AI

Website analyzing a reciept about busses
Inventory page with the ingredients and quantity and expiration dates
Main page with all the information

Inspiration

As University students who transitioned to living independently, one major issue we face is keeping track of the fresh ingredients in our fridges. During the first year of university, it was all too common of an occurrence where the fridge would have an "off smell", or our greens would be turning yellow, many times some produce that was fresh 2 days ago had turned into an inedible mess. This is not an issue that occurs on a small scale, in Canada more than 50 million tonnes is wasted every year [1]. Canadian households account for 47% of total food waste, and fruits and vegetables make up 45% of that waste [1]. This is why we decided to help this cause with our website PantryLens AI.

What it does

Our website helps people track the quantity of ingredients from a receipt and estimates an expiration date. This feature enables users to quickly record their purchases, alleviating stress from having to note it down manually. This also helps to minimize food waste as the users are able to keep track of what is about to go bad. Although it may not be 100% accurate, it can still help people be mindful of what ingredients to use up quickly.

How we built it

One of our main functionalities is to analyze images and send it to cloud vision to extract all the data we need from it. This information is then sent to Gemini AI to organize the data based on name, quantity, and expiration date. This information is processed and shown on the main page of the website. Our website also allows you to input unique ingredients one by one. All of this information is sent to an inventory that allows you to search any ingredient you have and allows you to tell the website how much of each ingredient you have used. This updates the total quantity left. You can also clear any ingredient you do not want on the list.

Challenges we ran into

Since this is our first hackathon we had a lot of trouble communicating important information to keep the workflow going. At times we would talk about too much unnecessary information and others would have a hard time telling each other important information we need. It was also difficult figuring out how to implement the AI into the website that would work as we would like it to, especially since it was our first time working with AI.

Accomplishments that we're proud of

We are extremely proud of getting the AI to analyze the receipt and properly output the information. Moreover, the menu bar transitions smoothly, adding a polished, user-friendly touch to our website. Our website can also analyze any product type.

What we learned

We learned a lot of technical skills, we were able to personally witness how AI can be used in so many different ways.

What's next for PantryLens AI

We would like to polish our website and include a section with recipes. This section would allow the user to search up recipes they would like to make, and the website will recommend ingredients that the user is missing. We would also like for the website to recommend different recipes based on existing ingredients.