MVP for Tracking Calories and Activity with AI

82%

first-try task success (log first meal)

first-try task success (log first meal)

7

weeks to MVP (24 screens, 3 core flows)

weeks to MVP (24 screens, 3 core flows)

2

iterations per feature (median)

iterations per feature (median)

100%

AA contrast

AA contrast

≥ 44px

touch targets

touch targets

Overview

Designed and launched an AI-powered calorie and activity tracking app that helps users log meals by simply taking a photo. The system combines photo recognition with a token-based system to simplify nutrition tracking, making it easy for users to monitor calories, macros, and daily activity. Working directly with the client, I created a monetizable MVP on a tight budget, focusing on clarity, user flow, and differentiation from competitors.

[Client]

AppLine/ A.Schtelbe Technologies

[Industry]

Health & Fitness / AI

[Role]

Senior UX/UI Designer (MVP & Product Strategy)


[Year]

2023

Research & Opportunity

The opportunity

Instant photo recognition, explainable results, and lightweight onboarding on mobile.

The goal

Reduce effort to the first successful log, keep AI transparent and correctable, and introduce clear, unobtrusive monetization.

Technical Constraints (framing the solution)

Low-latency upload/inference, fixed taxonomy (~70 classes), human-in-the-loop corrections, progressive permissions, GDPR, and encrypted upload.

UX Strategy

Three core flows:

  1. Snap → calories (camera-first, instant estimate)

Users snap a photo, the AI suggests what’s in the meal, and shows the estimated kcal count.

  1. Refine Predictions

Users can tap and edit the AI suggestions — which feeds future accuracy and improves the model.

  1. Token system

Each AI scan costs 1 token. Users get a daily amount, and can earn more by watching short ads.

This added a light gamification layer without pushing users into subscriptions.

Early Testing & Iterations

Token clarity: counters, just-in-time tooltips, subtle spend animation.

Navigation load: from crowded tabs → clean menu + floating action.

Final Design & Features

AI-Powered Meal Recognition

Users take a picture of their food, and the app auto-calculates the calorie content

Manual Adjustment of AI

Users can correct AI predictions and improve future accuracy.

Ad-Supported Model

Users can view ads in exchange for tokens, integrating monetization without disrupting user flow.

Outcome

A clear, explainable AI tracker that delivers photo in → calories out with transparent economics and a foundation for growth.

Research & Opportunity

The opportunity

Instant photo recognition, explainable results, and lightweight onboarding on mobile.

The goal

Reduce effort to the first successful log, keep AI transparent and correctable, and introduce clear, unobtrusive monetization.

Technical Constraints (framing the solution)

Low-latency upload/inference, fixed taxonomy (~70 classes), human-in-the-loop corrections, progressive permissions, GDPR, and encrypted upload.

UX Strategy

Three core flows:

  1. Snap → calories (camera-first, instant estimate)

Users snap a photo, the AI suggests what’s in the meal, and shows the estimated kcal count.

  1. Refine Predictions

Users can tap and edit the AI suggestions — which feeds future accuracy and improves the model.

  1. Token system

Each AI scan costs 1 token. Users get a daily amount, and can earn more by watching short ads.

This added a light gamification layer without pushing users into subscriptions.

Final Design & Features

AI-Powered Meal Recognition

Users take a picture of their food, and the app auto-calculates the calorie content

Manual Adjustment of AI

Users can correct AI predictions and improve future accuracy.

Ad-Supported Model

Users can view ads in exchange for tokens, integrating monetization without disrupting user flow.

Outcome

A clear, explainable AI tracker that delivers photo in → calories out with transparent economics and a foundation for growth.

Next Project

Next Project

Next Project

Responsive File Uploader for Blog Editors

2024

Let's Connect

If you’re looking for a designer to join your team or lead your next product initiative, I’m open to new opportunities. Whether it’s a full-time role, freelance collaboration, or just a conversation about how I can help, I’d love to hear from you.

Contact

Let's Connect

If you’re looking for a designer to join your team or lead your next product initiative, I’m open to new opportunities. Whether it’s a full-time role, freelance collaboration, or just a conversation about how I can help, I’d love to hear from you.

Contact

Let's Connect

If you’re looking for a designer to join your team or lead your next product initiative, I’m open to new opportunities. Whether it’s a full-time role, freelance collaboration, or just a conversation about how I can help, I’d love to hear from you.

Contact