Joblessness is skyrocketing and tens of millions of people are thrown into economic uncertainty. It is causing sprawling lines at food banks as people desperately seek their next meal.

According to University of Manchester Research, people say there is always five lots of stuff left over and it stresses them when the food is wasted. 60% of Canadians throwaways were edible, costing more than $1,100 per year. An average American spends over $7500 over food, but less than 30% is from home cooking.

To solve this issue and reduce food wastage, we have launched our AI chat web application Peter Pan, which provides you easy and organized way to create a mouth-watering recipe sand enhances your outdoor experience.

You can simply create your meal during outdoor activities or camping in minutes with the web application. Hiking, camping, no matter what activity you are doing, do not worry about what to eat when planning your next trip. Peter Pan is made to be your best travel companion no matter where you located in the world.

What it does

Peter Pan is a free software web conversational AI application that recommends dishes based on the ingredients you have, and tells you the recipe step by step. It is not only the best quarantine cooking sidekick, but also the best travel cooking companion no matter where you are located in the world.

It gives you three options for inputting your ingredients you have in hand – take a picture, input with text, or simply talk to it, then this app will offer a wide variety of tasty recipes.

At first, Peter Pan asks the user to input name, ingredients you don’t want, and health concerns in the profile section before it starts listening to the available ingredients. After customizing the options, you can take a picture of the ingredients you want to use cooking. You have to confirm which one to pick from two options for the meal convenience checkup. The quick option provides menu with least number of steps, and the creative option provides menu with the greatest number of steps. Just tell Peter Pan what you feel like having.

After that, Peter pan will recommend some menus according to its selection algorithm and your preference. Don’t worry if you don’t feel like taking the menu we recommend because we are going to provide other ones until you are satisfied. If you feel like you are behind following the steps of the recipe, don’t hesitate to tell Peter Pan. You can always ask Peter Pan to repeat the steps.

Once you are done with following the steps, smart Peter Pan will calculate and tell you how much you have saved by cooking yourself today. You can add the result to Google calendar and track how much you have saved by cooking.

How I built it

Website: Html, CSS, javascript, Figma Computer vision: Python (Google computer vision API) Back-end: node.js, Flask (Google cloud platform) Conversational machine: voiceflow (Google Assistance), recipe extraction (Spoonacular API)

Challenges I ran into

Trying to publish a fully functional and a visually compelling application on time was very challenging. We have all came up with a bunch of features we want to implement and each of our members worked on different parts of the application such as front-end, back-end, Voiceflow, and image recognition, so we had a hard time trying to put the all the parts and functions together.

When we started working on this project, our team was not very familiar with using Google Cloud, Voiceflow, and Google Cloud Vision API, so we had to spend some time familiarizing ourselves getting used to its features. One of the most challenging tasks was publishing the application using Voiceflow and integrating it with Google Assistance.

Even though we had to ask our mentors and spend much time on getting confident with the developing tools, we are very proud that we have but we have ended up understanding and published a conversational web application.

Accomplishments that I'm proud of

We are very proud that we were able to produce a fully functional and a visually compelling app. According to MLH regulation, we are allowed to develop and continue on the project idea we had previously. When we first came up with this project idea, we were not able to publish our application from Voiceflow nor connect our application with Google assistance.

We had an idea of getting the ingredients using computer vision, but we had to eliminate the function since it was hard to connect the image recognition codes to the back end. Successfully adding the blacklist section for the image recognition allowed Peter Pan to filter out more odd results such as paper effectively. Now, Peter Pan is ready to recognize whatever ingredients you have in hand.

Even though we have developed from our previous idea, we have come up with a completely new codes – from front-end structure to website. We think the app shows that we have spent much time on working on the UI part and customizing the recipe recommendation algorithm as well. Peter Pan is very user friendly and customizable – it almost feels like you are having a conversation with your friend what to eat for the upcoming meal.

Peter Pan allows you to save your recipes and export how much you have saved from cooking to Google Calendar. I think we have successfully managed the economic problem we wanted to solve from the first stage of the development.

We were very efficient in working since we have divided up our works from the beginning; we knew our strengths from the start. All of our members actively expressed opinions and we were very open to learning new developing tools and ended up developing a completely free and ad-free application that helps you save your money, time, and food wastage efficiently.

What I learned

Since we had a limited amount of time to finish our project, it was significant to plan the dues for each feature realistically. I think we have done a good job of using the Google APIs that expedited our process of development.

It was our first time using Mongo DB, Google Assistance, and exporting data to Google calendar, but we have successfully added all the features we have planned and published the web application.

What's next for Peter Pan

It would be a dream for us to further polish the app and deploy it to app stores. We are very contented that we could add various functions that help customize and recommend the best recipe for a user, but it would be better if we could make Peter Pan work offline as well. If we were given more time, we would like to implement some conversational features that cheer people up during the pandemic and further expend this conversational AI into other fields.

Built With

Share this project: