Agentic AI Assistant for the Modern Chef
Version 16 (Culv16.py) picks a popular recipe from the selected geographical region (goes through the 50 states of the USA). After that, it generates the recipe, as well as the costing of the top ingredients (by price). The LLM then generates a description of an elegant presentation, including plating instructions and other food items that are often served alongside this dish. Finally, a Diffusion model creates an image of the recipe from the LLM generated description. This repeats forever until stopped - thereby creating a vibrant slideshow of food photos, including state, recipe name, and ingredient costs.
Background
We had a great time at out "Chef's Buddy" AI seminar at the Richmond, VA campus of ECPI University back in February of 2025. At the seminar, non-engineers (Culinary Students) saw how a first version of the "Chef's Buddy" AI Agent works. The culinary students also got a chance to play around with popular free online LLM's.
Alongside the culinary students were Technology students (studying electronics engineering technology and other subjects). The technology students had to gather further requirements by listening carefully to the ideas of the culinary students. We hope to use these requirements to make improvements on the first version of Chef's Buddy.
This repository is a space where that work can take place. The original source code for Chef's Buddy is located here, along with the gathered requirements.
We hope to use this repository to teach how to use Github, Agentic AI, and also light requirements tracking. All are welcome to participate!
Want to learn more and see pictures from the event? Here is a LinkedIn article about the event:
Here is a LinkedIn article about the technical details of the implementation:
https://www.linkedin.com/pulse/chefs-buddy-technical-details-paul-nussbaum-twmne/