RecipAI

Inspiration

It started when I opened my uni fridge one morning and checked my cupboard. An onion, some salsa, some spices and cheese. I had no clue what I could do with this random assortment of items or how to begin creating a meal. If only there was a tool to help me use up the rest of my ingredients to prevent them being thrown away and increasing food wastage.

What it does

The site has a database with four main tables, users, recipes, ratings and pantries. These each hold data used within the site. Users for login data. Recipes for storing the recipes and their average ratings. Ratings for storing the users ratings of recipes. Pantries to store the users pantry. Within the sites then gemini is linked with an API key in order to pass commands with the ingredients and dietary preferences. The community page works by displaying all of the recipes that have been generated and is sorted by ratings. The cookbook shows the user the recipes that they have generated.

How we built it

We started the process by figuring out what the main functionality of it would be, we got the generating recipes from ingredients and an image of a meal and a login system. These systems we were quite quickly able to implement, after completing these then we got the CSS set up to make sure that the UX was as easy and seamless as possible. Once this basic functionality was in place then we started brainstorming additional features. We came up with a pantry to store the users ingredients, a cookbook of their previous recipes, a rating system and a community page where all the recipes were stored.

Challenges we ran into

For most of the group it was our first time working as a group on a single project so managing to split up tasks so that everyone had something to do was a challenge at first but we managed to overcome this as time went on as we improved our communication skills and ability to read each others code. Moreover, it was our first time using a GEN AI API key and with that came many challenges such as making sure prompts were structed correctly and that the AI doesn't break or ignore rules placed upon it.

Accomplishments that we're proud of

We are proud of how seamless the webpage looks despite the limited time we were given. We wanted to give the user a great experience and we believe we acheived that. We also managed to branch off from the original idea and think of lots of great add ons to help enhance the project.

What we learned

Ultimately, we learned a plethra of skills and it was a invaluable experience to us. One of the most important things we learnt was resilience to constantly stay on task and working for most of the 24 hours. Other things we learnt was how to read other peoples code and how most people don't code the same so we had to adapt constatntly throught the event.

What's next for RecipAI

There are many potential future paths for RecipAI. One such being integrating it into other devices for example a smart fridge could have an inbuild link to the site and use the camera to update their pantry. There is also potential for affiliation with shops that are promoted on the site to help stock up the users pantry. Subscription models could also be implemented in order to fund further investment into the site for upgrades such as meal prepping system or creating a profile of the user to further specialise the recipes for them.