CarePlate

Inspiration

We started with a simple observation: eating safely isn’t simple for everyone.

For many older adults, meals aren’t just about preference; they’re tied to medications, chronic conditions, and strict dietary limits. Something as small as the wrong ingredient can unintentionally trigger serious consequences like worsening chronic conditions or unsafe medication interactions, but keeping track of all those factors every day is overwhelming.

We wanted to build something that could take that mental load off. Instead of expecting users to remember every restriction, CarePlate helps guide their choices in the moment.

What It Does

CarePlate is a smart assistant that helps users decide what they can safely eat.

The large language model takes into account medications, health conditions, and dietary restrictions, then generates meal suggestions that fit those constraints. Users can check foods, get meal ideas, or even chat with the interactive AI assistant, all without needing to understand the underlying complexity.

The focus isn’t solely convenience, but also reducing risk and helping users feel more confident in their daily decisions.

Different Features:

• Daily tracker : Tracks meals, water, and medications
• Suggested recipes : Personalized depending on your medical condition, dietary/religious restrictions, or even based on cuisines, seasons, and occasions
• Meal planner : Plan out your daily meals
• Grocery list : Grocery list that has all the items from the recipes you are planning on making. Has the costs of each item depending on what grocery store you shop at. Has a built in delivery function or a function that sends the list to the caretaker.
• Food risk check : Assesses your medical conditions and rates foods specific for you specifically.
• Interactive AI assistant : Can talk or chat with the assistant, asking questions like "What should I eat today?" or "Is this food safe to eat with my medication?"
• Caregiver check-in : There is a caregiver view that allows them to see if they ate their meals and took their medications without having to constantly check in with them. Alerts are sent if the medication is missed.
• Personalized profile : Customizable profile depending on dietary and religious restrictions, medical conditions, and emergency contact information.

What We Learned

Through building CarePlate, we learned how to design technology that is both accessible and practical for seniors. We gained experience integrating user-centered design principles, such as large, intuitive buttons and simple navigation, while also incorporating AI to provide personalized food recommendations. We also learned about the complexity of medication–food interactions and how important it is to account for real-world health risks when creating a health-focused application.

How We Built It

We designed CarePlate to combine flexible AI responses with structured safety checks.

It is a language model that handles user interaction, paired with a rule-based system that filters and validates outputs against known constraints like food–drug interactions and dietary limits. This prevents unsafe or misleading suggestions while still allowing the system to feel conversational.

By sending in studies, the model was able to have a better understanding of food-medication based interactions, dietary restrictions based on medical conditions, and more.

We also prioritized accessibility on the interface side, keeping interactions simple and clear so the tool is usable without a steep learning curve.

Challenges We Faced

One of the biggest challenges was making sure the system could be both helpful and reliable.

AI naturally generates suggestions, but in a health-related setting, accuracy matters much more than creativity. We had to rethink how outputs were generated and add safeguards to ensure recommendations stayed within safe boundaries.

Another challenge was designing for a wide range of users with different needs and levels of comfort with technology, which pushed us to simplify the experience as much as possible.

Looking Forward

Looking ahead, we want CarePlate to move from being a helpful assistant to something people can rely on every day.

One direction is integrating real-time health data, so recommendations can adjust based on things like blood sugar levels or day-to-day changes in condition, instead of staying static. We also want to strengthen how the system anticipates risk, not just reacting to obvious conflicts, but catching smaller issues before they become problems.

Built With

• claude