Design Leadership

AI-Powered Nutrition Feature

AI-Powered Nutrition Feature

Helping diabetes patients understand their food choices through intelligent meal analysis

 

 

The Problem

Nutrition tracking was a major barrier for diabetes management. Our research with 6 diabetes patients revealed:

  • Tracking was too manual: "I wish I could just say what I ate instead of typing it all in"

  • No personalized guidance: Patients were trying multiple apps, finding none that fit

  • Knowledge gaps: Patients didn't understand how food choices impacted their diabetes

Core insight: To manage diabetes effectively, patients need to understand the food they consume.

 

 

The Solution

I designed a multi-modal AI nutrition feature with three input methods and intelligent feedback:

Key Features

  • Photo, voice, or manual entry - meet patients where they are

  • AI analysis - instant nutritional assessment for every meal

  • A-C grading - simple scoring for immediate understanding

  • Constructive feedback - positive reinforcement + actionable suggestions

[VISUAL: Meal summary page showing grading and feedback]

 

 

Design Philosophy

Working with a nutritionist, I created a feedback framework that:

  • Highlights positives first ("Keep doing")

  • Provides diabetes-specific suggestions

  • Never judges food choices

  • Encourages incremental improvements

 

 

Implementation & Results

Strong Engagement

  • 66% adoption rate among patients

  • Weekly logging consistency maintained

  • Manual entry most popular despite photo/voice options

 

This nutrition feature represents the first step in Nuna's evolution toward AI-native healthcare experiences. The success validated our approach to combining clinical expertise with intelligent technology, setting the stage for our broader voice AI assistant and future AI-defined interfaces.