Breader Together

As new adults in university, our team found it difficult to live away from home and balance the difficult art of culinary creation with the primal necessity of consuming calories. Breader Together fuses appetizing social media, a personalized recipe provider and gamified cooking challenges. The result of engaged individuals is the jumpstarting of a global social media platform of endless creation, community-driven sharing and gastronomical inspiration for all, open to all.

Powered by Cohere’s powerful in-house (Command) LLM models, an endless buffet of delicious dishes is only a click away. And maybe a few stirs, of course 😎. Using Cohere’s LLMs, we can personalize any recipe based on what ingredients you have, cooking experience, time constraints, or if you want to simply try something new!

Now that you’ve whipped up a mouth-watering meal, it’s time to share it with your friends and the world! Post the meals you’ve made on the feed, and also get inspired by a global community of home cooks. Don’t forget to “loave” 🧡 your favourite posts and 🌮 taco bout it in the comments!

Ready, Set, Dough! Looking to level up your cooking skills? Batter up at our fresh and exciting challenges, to cook certain dishes, achieve healthy eating goals or simply to engage with other home chefs for inspiration. Gamify your culinary journey and become a pro in no time!

How We Built ‘Breader Together’

Breader Together uses a depth of different technologies and frameworks. The core of our project is powered by Cohere’s text generation API which is used to provide highly-personalized and accurate recipe recommendations to our users. Our full-stack application also has a decoupled front/back-end and a SQL database used to store our data. Our website’s frontend (UI) is designed using React with Tailwind CSS and Axios (HTTP Client) to create a comprehensive (featured) and ergonomic user experience. On the other hand, we used FastAPI to rapidly prototype a functional API microservice that could be deployed using container technologies such as Docker.

Challenges We Faced

We initially had quite “over-engineered” solutions to approach the challenge we were trying to solve. Our initial idea involved many moving components and had many constraints which we were not satisfied with. However, we realized that Cohere’s LLMs had many desirable features that not only make it an ideal LLM, but in-fact the perfect solution for our use case. However, after experimentation, we settled on using Cohere to generate recipes, while accounding. Lastly, as our team used Axios and FastAPI, we found some unexpected (ad-hoc) incompatibilities while integrating our full-stack application, while parsing user input during recipe generation.

What’s Next for Breader Together?

What’s next? That’s a big question. Saying goodbye could be sad, so maybe we can chat about it over a nice warm bowl of chicken noodle soup, while reading some books. Just kidding. Like any cook, we love to create, bring joy and foster a community with our creation. As such, we would be extremely eager to cook up new features and keep our app running with the help of sponsors (like Cohere). We believe that our vision could develop into something magical for cooks and cooks-to-be around the world.

Peace out and enjoy your next croissant. Or pad thai. Or reverse seared steak. It’s almost if the possibilities are endless. Wait. Does that remind you of a certain app?