Project Story: Cocktail Creator

Inspiration

The inspiration for the Cocktail Creator project came from a shared passion for mixology and AI technology among our team members. We wanted to explore how generative AI models could be used to create personalized cocktail recipes based on user input.

What We Learned

Throughout the development process, we learned a great deal about prompt engineering and foundational models (FMs) in the context of generative AI. We discovered how to fine-tune models to generate specific types of content, in this case, cocktail recipes. Additionally, we gained insights into user experience design and the importance of creating an intuitive interface for interacting with the AI-generated content.

Building the Project

The project was built using PartyRock, an Amazon Bedrock Playground, which provided a fast and fun way to experiment with generative AI. We utilized pre-trained models and customized prompts to generate cocktail recipes based on user-provided keywords or ingredients. The backend was implemented using AWS services for hosting and managing the AI models, while the frontend was developed using HTML, CSS, and JavaScript for creating an interactive user interface.

Challenges Faced

One of the main challenges we faced was fine-tuning the AI models to generate accurate and coherent cocktail recipes. This required extensive experimentation with different prompts and model configurations to achieve the desired results. Additionally, integrating the backend AI services with the frontend user interface posed some technical challenges, but we were able to overcome them through collaboration and problem-solving.

Built With

• python