Targeted Tracks → Green House - Best Use of Cohere API

Inspiration

SnapNCook was inspired by the urgent need to address food waste and promote sustainability in our daily lives. We saw an opportunity to tackle food waste while helping people get creative in the kitchen. Our mission is to empower households to reduce food waste, save money, and inspire culinary creativity with AI-driven meal solutions. By minimizing food waste and encouraging resourceful meal planning, we aspire to foster eco-friendly living and contribute positively to global sustainability efforts.

As a passionate advocate for sustainability, I believe that immediate action is crucial to protect our planet for the generations to come. We need to promote sustainability in a way that is easy, fulfilling, and fun for people involved. By reducing food waste and inspiring creative culinary endeavors, we strive to contribute to a more sustainable and environmentally conscious future, one meal at a time.

What it does

SnapNCook is a smart webapp designed to promote sustainability by helping people make the most of the ingredients they already have. The goal of the app is to make sustainability beneficial for everyday people. Studies have found that we waste about 1.4 billion tons of food every year. Interestingly more than 80 percent of Americans discard perfectly good food because they misunderstand expiration labels. Understandably, in an effort to avoid food-borne illness, people will throw out good food. We found we can make a huge dent in the food waste problem by targeting households and encouraging the use of food that is still good but will soon expire.

Although recipe creator apps already exist and are successful, they don't encourage users to be sustainable and they don't help users save money. That's where SnapNCook comes in. We aim to create a sustainable solution that allows users to use all the food in their fridge, reducing food waste and saving users money.

UX FLOW

1. Take a picture of your fridge
2. Identify items and expiry dates
3. Provide recipes based on foods expiring first

How we built it

We had fun building a full-stack application utilizing machine learning technology to solve a real world solution.

Front-end: Javascript, Next.js with Tailwind CSS Backend: Flask, Python ML Stack: Cohere API, Python, Flask, YOLO, TensorFlow, Google Cloud Vision

Challenges we ran into

One challenge we ran into was getting a working object detection model. It took us a little bit to figure out how to go about implimenting this and then including it into our backend. But nothing an all-nighter couldn't fix!

Accomplishments that we're proud of

We are proud of finishing our project on time with the main features we set out to build. We had a lot of fun staying up late hacking!

Completing this project is a big accomplishment for us. We spent a lot of time formulating an idea and researching the way we wanted to help make the world a more sustainable place.

What we learned

We learned a lot about implementing machine learning models into real world applications.

What's next for SnapNCook

Design wise, we would look to make the app responsive so it can be used on a wide variety of mobile devices. We would also look to improve accuracy in our object detection model by allowing it to use more data as inputs. Additionally we would teach the model to recognize more foods. We would also look to improve accuracy in the classification of our object's expiry dates. We would again provide the model with more data and train it to recognize food going bad and assign that with an expiry date.

Built With

• cohere
• flask
• google-cloud
• javascript
• next
• python
• react
• tailwind.css
• tensorflow
• yolo