top of page


Skills Utilized

2 weeks, November 2022

User Research
User Interview
Visual Design

“Sous-Chef” is a mobile app that helps improve users' cooking experience and diets. It is an extension of the features mentioned in "Architecture Machine" by Nicholas Negroponte. It uses ML and AI to create a highly personalized cooking experience and diet plan for each user.

01 Historical concet

Nicholas Negroponte founded in 1967, together with Leon Groisser, the Architecture Machine Group at MIT. Negroponte’s vision was an architecture machine that would turn the design process into an altering of traditional human-machine dynamics.

His approach was significantly influenced by a recent discussion on:

  • Artificial intelligence

  • Cybernetics

  • Conversation theory

  • Technologies for learning

  • Sketch recognition

  • Representation

Particular subassemblies that should be part of an architecture machine:

  • Heuristic mechanism

  • Rote apparatus

  • Conditioning device

  • Reward selector

  • Forgetting convenience

02 Scope and intentions

The Architecture Machine is not a fully developed machine that holds the ability to do certain tasks; rather, it’s more of a directional concept for the future development of machines. That being said, there are a lot of possibilities/ flexibility to where this theory could be applied.

I propose to prototype a cooking-related technology that reflects the core value of the architecture machine.


The target audience of this app is mainly college students and office workers who want to improve their diets and cooking skills. Their proficiency in cooking could vary and the objective is to establish a fun,  convenient, and meaningful experience through cooking.

I asked the potential users about their most commonly encountered issues in the process of cooking. 


04 Visual identity

The modern and clean design of the cooking app is characterized by warm hues of mustard yellow and brown, conveying a sense of friendliness and reliability to users.

UI KIT.png

05 digital Prototypes

portfolio 1.png

Sign In Page 

portfolio 2-1.png

Onboarding Questionnaire: Dietary Restricitons, Cooking frequency, Proficiency, Goals ...

portfolio 2.png

User Profile Page |  Expand Menu

portfolio 3.png

Connect to your local grocery store for an easier shopping experience: users can log in and connect to their Giant Eagle account and use their “Scan, Pay, Go” to automatically update their grocery list and new purchases. Connect to the smart fridge: with camera setups and AI recognition, the fridge will be able to tell users about in-stock items. Sous-chef allows the users to check the items and generate recommendations based on the available ingredients.

portfolio 5.png

Key Feature:  Voice-based chatting system with the sous-chef. It allows the user to have a conversation before, during, and after cooking.

Smart Recommendations
Example: suggests recipes under the desired cooking time limit and style of cuisine.

Problem Solving
Example: How to improvise when running out of certain ingredients.

Chef & Sous-chef Collaboration
Example: creates new personal recipes based on pre-existing recipes.

portfolio 6.png
  • In “My Cooking Book” is a collection of recipes/ dishes made by the user. It’s a way to keep track of progress and history.

  • Each recipe could have its instructions played out loud on the speaker, providing the user with a hands-free experience.

  • Each recipe would have a rating from the user. The rating also influences the chance of it (or dishes similar to it) showing up in future recommendations.

06 Conclusion

The architecture machine theory provides a flexible framework for understanding and applying concepts related to machine-human interactions. For this project, I decided to shift my focus from art and design to explore other domains, such as cooking, where these concepts can be applied. In particular, I wanted to investigate how to facilitate effective communication between machines and humans.


Through my research, I learned that this type of communication should be highly personalized and able to introduce new ideas, but not so complex as to be incomprehensible. The goal is to create a balanced and mutual understanding between humans and machines. My prototype reflects this approach, with input from both machine and human sources.

portfolio 1.png
bottom of page