Snapcarbs, an app that monitors your carb intake.

Product Overview

This product is designed to help people with diabetes and general users monitor their carbohydrate intake. By taking photos and uploading food images to the database, cloud-connected nutritionists can analyze the food content in real time, calculate the amount of carbohydrates, and record and track daily intake.

The product also provides calorie calculation and weight-tracking features to help users better manage their overall health and nutrition.

Team 

2 App Engineers • 1 UI/UX Designer (Me) • 1 Nutritionist • 1 Product Manager

Interface

Mobile App (iOS/Android)

Development Time

12 weeks

Design Objectives

Simplify carbohydrate tracking to improve usability.

Design Challenges

People with diabetes often lack professional training, making it difficult to accurately estimate carbohydrate intake.

• Maintain scalability for future expansion into additional nutritional metrics, such as calorie, fat, and other dietary values.

• Food types vary significantly, creating challenges in consistent identification and estimation.

• Serving containers and portion come in different shapes and sizes, making portion size estimation difficult.

• Users’ photography conditions vary widely. Factors such as focus, lighting, camera angle, and ambient color can affect the accuracy of food recognition and nutritional estimation.

Success Metrics

8/10 testers reported that the app does help them to track the intake of carbs every meal, making the monitoring diabete easier.

Development Process

Discover

Explored real user issues and observed limitations of current technology.

Define

Identified core problems and formulated the optimal product solution.

Develop

Collaborated closely with engineers, dieticians, and diabetic patients to refine and build the product.

Deliver

Successfully drew the intentions from investors and startups.

Background

The key to diabetes glucose management lies in consistent monitoring, understanding target ranges, and maintaining accurate records. Patients should follow medical guidance and measure their blood glucose levels regularly. The recommended targets are to maintain pre-meal blood glucose between 80–130 mg/dL, keep post-meal blood glucose below 160 mg/dL two hours after eating, and maintain glycated hemoglobin (HbA1c) levels below 7%.

User Interviews

I conducted interviews with 2 categories: diabetic patients and medical professionals:

After the user interviews, I summarized the key pain points and advantages:

Pain Points

• Difficulty estimating carb intake, especially when eating out or consuming mixed dishes.

• Manual food logging is time-consuming and easy to forget.

• Lack of accurate carb information in restaurants, delivery apps, and school cafeterias.

• Irregular lifestyles and busy schedules reduce consistency in tracking.

• Social events, snacks, and cravings make diet control difficult.

• Elderly users may struggle with small text and complicated interfaces.

• Existing apps often lack support for Asian foods or personalized recommendations.

• Users want faster and smarter tracking methods instead of manual input.

Advantages / Opportunities

• AI food recognition can simplify carb tracking through photos.

• Integration with CGM, smartwatches, and health apps improves monitoring accuracy.

• Voice input and automated tracking reduce user effort.

• Personalized recommendations can help users choose healthier meals.

• Gamification and reminders can improve engagement and consistency.

• Community food databases can provide real-world carb information.

• User-friendly accessibility features benefit elderly patients.

• Real-time alerts help users prevent excessive carb intake and maintain better glucose control.

Mockup User Flow

Result

Improved usability and accuracy: The app significantly improves both ease of use and carbohydrate estimation accuracy. Unlike traditional methods that require users to visually estimate portion sizes and manually calculate carbohydrate content, the app automates the process using AI image recognition.

Designed for everyday users: Most people are not registered dietitians or nutrition experts, so estimating carbohydrate intake by eye often results in large errors. The app reduces the knowledge barrier by providing instant and consistent carbohydrate estimates.

Fast and convenient: Users simply take a photo of their meal, and the app automatically identifies the food, estimates the carbohydrate content, and records the meal in the app. This greatly reduces the time and effort required for daily food logging.

Encourages consistent tracking: By minimizing manual input, the app makes it easier for users to develop a long-term habit of monitoring their carbohydrate intake, which is essential for effective diabetes management.

Future opportunities: Apple's upcoming enhanced depth and distance sensing could potentially be integrated to estimate food volume more precisely. Combining these technologies with AI image recognition may significantly improve portion size estimation and overall carbohydrate calculation accuracy in future versions of the app.