UX Research & Design
School Project

AI-powered application to assist people to maintain a balanced diet and achieve health goals

Project Philosophy

Technology can help reduce people's cognitive load, if well-designed.


Proper nutrition intake is essential not just for physical health but emotional and mental care. There are many well-designed applications in the market to track nutrition consumption and give diet suggestions. Yet most people fail to continue using them. We decided to explore the key issues young people face in this regard and design a solution.

On the whole, what differentiates our application from many other nutrition applications is that we provide a holistic, one-stop personalized interface for people to manage their food, find help in maintaining a diet, as well as track their progress while staying motivated.


  • Aug 2021 - Dec 2021


  • Huixiang Chen
  • Apoorva Gondimalla
  • Mayur Salunke
  • Bo Zhang
  • Jiaxin Zhang

My Role

  • UX Researcher
  • UX Designer


  • Figma
  • UserTesting

Problem Statement

"How might we reduce the abandon rate if tracking is such a tedious and annoying experience?"

Design Process

Exploratory Research
In-Depth Research
Exploratory Research
In-Depth Research

Exploratory Research

Competitive Analysis

I did one round of competitive analysis to understand our competitor's strengths and weaknesses to find a gap in the market.

Table of main functions of 5 competitors, categorized into three groups: motivation, support, and data input processTable of main functions of 5 competitors, categorized into three groups: motivation, support, and data input process
Problem Space
  • How might we help users stay motivated?
  • How might we provide support?
  • How might we design for a less tedious data-inputting experience?

In-Depth Research

41 surveys
To find patterns
  • We launched a survey to investigate people’s diet-related habits and their perspectives on diet regime
  • The survey includes 15 questions separated into 2 categories, demographic/background questions and diet-related behavioral questions
8 interviews
To understand the reasons
  • We conducted semi-structured interviews to identify specific user challenges surrounding food consumption
  • The questions included in the script are mainly grand-tour questions and a set of mini-tour questions related to each grand-tour question, which can help avoid leading questions and truly understand users’ behaviors


We did our data analysis using an affinity diagram. Four major categories of pain points were identified.

Bo and her teammates making an affinity diagram in a group study room


Time and Effort
  • People usually skip meals or change their meal plans due to busy schedules.
  • People tend to eat out, order food online, or cook simple dishes because of their tight schedules.
  • People stop nutrition tracking because they don’t want to put in the effort to manually log food and find it mundane.
Recipe Suggestions
  • Some people prefer videos with quick instructions, like TikTok, while others prefer detailed instructions.
  • People are unsure if they are approximating the right quantity while cooking and tracking.
  • People cook what they already know in detail repeatedly.
  • People want to check if they have the ingredients first.
  • People are looking for new and healthy recipes to achieve their goals.
  • People are unable to make quick decisions since there are too many different information sources for recipes and instructions.
  • People have to consider the expiration period of each food to avoid food wastage however it’s difficult to keep track of everything.
  • People don't really know the nutritional value of food items they consume when they eat out.
  • People have difficulty reading and understanding nutrition facts.
  • People often give in to cravings.
  • People are unsure if maintaining a balanced nutrition diet is worth the effort.

We created these three personas based on the four categories of pain points we identified to help us gain a better understanding of user needs by putting them into a specific context. These three personas would be the main artifacts we referred to in the ideation phase.


We conducted a brainstorming session, with the status quo of the current market and user needs in mind, to ideate the necessary features and solutions to solve the issues that we are currently experiencing.


  • Ethnicity filter for recipe search
  • Feedback based recipe suggestions with AI

Decision Making

  • Feedback based recipe suggestions with AI

Food Management

  • Track your pantry to prevent foodwastage
  • Item based recipe search

Food Tracking

  • Assist in food detection
  • Multi format input

Customized Notification

  • Connect with users’ calendar to plan food
  • Use location and time to remind users on their goals

Cooking Instruction

  • Multi format recipe instructions to accommodate different experience level cooking


  • Reward and involve users by gamification
  • Minimal, friendly

Design & Evaluation


Since there are a lot of information going on in our app, we started with a sitemap to capture all the major functionalities we designed.

Mid-Fi Prototype

We designed all the major screens of the app based on our sitemap.

Screenshots of grayish main screen mockups created using Figma with more structured information architecture and realistic languageScreenshots of grayish main screen mockups created using Figma with more structured information architecture and realistic language

Usability Testing




Think aloud

Follow-up questions

  • Some confusion was created by improperly used visual and textual cues that were introduced to indicate the function of the components and what has been done by the user.
  • Some confusion was created by redundant processes.

Hi-Fi Prototype

Some changes were made to alleviate the confusion that arose from the last round of testing. We also added colors, real photos, and video models to help us test how the user experience of our design was in a natural setting.

Screenshots of more realistic main screen prototypes created using Figma with colors and images addedScreenshots of more realistic main screen prototypes created using Figma with colors and images added
Major Changes
  • Removed the ‘Explore recipes’ button and revised the language in the custom notification as they make the users confused, they tried to click on that button to search for recipes, but what we intended to do is to give them the option to explore recipes cooked with that specific kind of ingredient
  • Gave each icon a name to give users an idea of what each icon stands for
  • Moved quantity estimation information next to the size-changing section making it more intuitive for users to get the idea that we are using the 'fist' concept to help users estimate food quantities
  • Replaced the too-detailed description of quantity left in ingredients with whether items are present or not, like ‘Available / Unavailable in your pantry’, to make users feel more intuitive about food quantity
  • Canceled the ‘grocery’ icon on the top so that users can search for the item directly on the same page for the add action
  • Added ‘My existing list’ to indicate that the food items listed below are what users have already
  • Replaced ‘good’, “average”, ‘bad’ with new terminology to describe food status
  • Redesigned the quantity information button to be more intuitive by making the entire information bar clickable and the window pops up when users try to add or subtract food quantities, instead of increasing quantities directly by clicking on the numbers on the information bar


Key Features

Onboarding Process

Understandable Activity Level

  • Options with example scenarios for users to better estimate their activity level

Accessible Amount Estimation

  • Calculating fist size to give custom estimation of food quantity indicated in the recipe or nutrition information

In-App Experience

Customized Notification

  • Notifications generated based on recent activities that lead the user to next ideal action to achieve their goals

Help in Recipe Decision Making

  • Limited personalized suggestions to quickly choose recipes to cook
  • Options to give feedback on options provided

Informative Pantry

  • List of items with information about when the food was bought, and its estimated condition

Seamless Experience Across Pantry and Recipe

  • Options to search for recipes based on the ingredients users have in their Pantry
  • Information on Recipe about availability of ingredients required in pantry

Lock Screen Notification

  • Notifications on the lock screen assisting users to plan cooking and meal times by synchronizing with their calendar

“Being able to search and save recipes based on how long they take to make, how they align with your health goals, and what ingredients you'd like to use is amazing. ”

Lessons Learned

Collaborate and Communicate

Each team member has a different educational background. This always adds a new perspective to the discussion table.

Be Flexible about Methods

There's such a huge UX toolbox. It's always good to pick the ones that best suit the context from it.

Optimize Content Strategy

Delivering both visual and text cues to the user in the right time, place, and format is really important to support a more seamless user experience.