Jul 06 '26
EatSmart CoChef: Hands-Free Food Intelligence for Smart Glasses
#AIFS
#Research
EDITOR’S NOTE
Most food technology promises efficiency. Eat this, avoid that, count these calories, scan that barcode. But for all the information available to us, making food decisions remains surprisingly difficult. We still stand in front of open refrigerators wondering what to cook, or scroll through recipes mid-cooking. We still struggle to connect nutrition advice with everyday eating.
At the UC Davis AI Institute for Next Generation Food Systems (AIFS), research engineers are approaching the problem differently. EatSmart CoChef, developed by research engineer Pranav Gupta, was demonstrated at AIFS's AI AgTech Challenges & Hackathons, showing how AI wearable technologies could deliver real-time food intelligence. Rather than asking people to stop and consult another screen, these wearables bring information into the exact moment a decision is made.
EatSmart CoChef combines computer vision, conversational AI, and food science data into a hands-free cooking assistant that can identify ingredients, recommend recipes, and explain the nutritional details of your food.
Q: What inspired the creation of EatSmart CoChef?
We wanted to remove the friction of getting real-time guidance while cooking, shopping, or exploring ingredients. Most food and nutrition apps require users to stop what they’re doing, pull out a phone, take a photo, type a question, and read a screen.
With CoChef, someone wearing smart glasses like the Meta Ray-Ban can simply look at food, ask a question, and receive spoken guidance. The goal is to make food intelligence accessible at the exact moment people make decisions.
Q: Can you walk us through a typical experience?
Imagine opening your refrigerator and asking, “What can I make that’s heart-healthy?” or “What should I cook with what’s on hand?”
The glasses capture a first-person view of the ingredients, and CoChef identifies what’s available. It connects that information with nutrition and food science data, including FoodAtlas and the Periodic Table of Food developed by UC Davis AI Institute For The Next Generation of Food Systems and Innovative Institute for Food and Health, then recommends recipes aligned with the user’s goals.
The experience remains entirely hands-free. Instructions come through the glasses, and users can ask follow-up questions, request substitutions, or learn why certain ingredients may support cardiovascular health, antioxidant intake, or anti-inflammatory eating. In that sense, CoChef behaves more like a knowledgeable cooking companion.
Q: How is CoChef different from simply asking ChatGPT for a recipe?
Food decisions are physical. When people cook, shop, or eat, their hands are occupied and their attention is elsewhere. Most AI recipe chatbots start with typing, which interrupts the experience.
CoChef begins with what you’re actually seeing. Users can look at food, ask questions, and receive guidance backed by FoodAtlas-powered intelligence without breaking their cooking flow.
This app is for anyone who has ever opened a pantry and wondered, “What should I make, and is it good for me?”
Q: Tell me more about FoodAtlas?
Recipe generation is only a small part of what CoChef does.
FoodAtlas provides the deeper food intelligence layer behind it, connecting ingredients with nutritional profiles, bioactive compounds, and wellness-related insights. That means users can learn not only what is in a dish, but also how particular ingredients may contribute to antioxidant properties, nutrient density, or other food characteristics.
The goal isn’t to provide medical advice. It’s to make food science more understandable and accessible during everyday decisions.
Q: How do you envision the future of AI-assisted eating?
We think food intelligence will become increasingly contextual.
Instead of manually logging meals, scanning labels, or searching nutrition databases, people will simply ask about the food in front of them. As wearable devices improve and food datasets become richer, AI systems will provide personalized guidance during cooking, grocery shopping, restaurant visits, and everyday meal planning.
The long-term opportunity is real-time support for making informed food decisions.
Try EatSmart CoChef
EatSmart CoChef is available today as an open-source iOS application for Meta Ray-Ban smart glasses. Explore the project, test the app, and share your feedback with the developers:
