Sustainaplate

• 

  Mockup of Sustainaplate

• 

  Mockup of Sustainaplate

• 

  Screens of Sustainaplate

Inspiration

Our inspiration for Sustainaplate stemmed from a shared passion for sustainability and a desire to make a meaningful impact. We've all found ourselves discarding food from our refrigerators because we didn't use it in time, and we can relate to the reluctance to check sustainability labels on products while shopping. These shared issues inspired us to create a tool that empowers users to make sustainable choices while cooking and shopping.

What it does

The Sustainaplate Backend serves as the heart of the Sustainaplate web application, orchestrating real-time recipe recommendations, ingredient analysis, and sustainability scoring. It seamlessly integrates with the front end to provide users with a dynamic and eco-conscious culinary experience.

How we built it

We crafted Sustainaplate as a web application primed for future expansion onto mobile platforms. Our technology stack hinges on a knowledge base index built using a state-of-the-art AI toolset - Vector Database with Retrieval Augmented Generation (RAG) method and stable diffusion for image generation. API served by FastAPI for the backend server while leveraging the power of React for the front end and integrating MaterialUI to guarantee a visually appealing and user-friendly interface consistent with Migros' branding. We engineered a bespoke API that seamlessly fetches and transmits recipe and sustainability data. Our approach involved meticulously examining the Migros-provided data and incorporating AI algorithms that Meta Llama2 Large-language-model (LLM) empowered to energize our recommendation engine. These algorithms adeptly assess user-inputted ingredients and deliver recipe recommendations, all while conscientiously considering sustainability aspects.

Challenges we ran into

Formidable challenges marked our journey. Effective integration of AI, especially for real-time recipe suggestions, demanded meticulous fine-tuning. Sourcing precise sustainability data and ensuring the reliability and adaptability of our sustainability scoring system across diverse user preferences presented complex intricacies. The optimization for mobile device compatibility and the unyielding pursuit of responsive design posed formidable technical hurdles. Nevertheless, these challenges served as the crucible that steeled our resolve to furnish a resilient, user-centric solution harmonizing with our overarching sustainability mission.

Accomplishments that we're proud of

We take pride in creating a comprehensive, user-centric solution that offers recipe recommendations and promotes sustainable cooking and shopping. Our achievement lies in harnessing AI to impact daily food choices positively.

What we learned

Throughout the development of Sustainaplate, we gained valuable insights into the potential of AI and technology to drive positive change. We deepened our understanding of the complexities of sustainability in the context of food, from ingredient sourcing to waste reduction. Additionally, we learned the importance of user-friendly design and accessibility, ensuring that individuals from all walks of life can embrace our platform.

What's next for Sustainaplate

We envision expanding Sustainaplate to include more advanced AI features like personalized meal planning, waste reduction strategies, and saving user preferences, and recommend substitutes for more sustainable food choices. We aim to continuously enhance the user experience and promote sustainability in every aspect of culinary life.

Built With

• fastapi
• google-cloud
• huggingface
• javascript
• jupiter
• lllama2
• materialui
• python
• rag
• react
• stablediffusion
• vectordb