UX/UI Case Study

AI-Driven Diet planner

This study demonstrates how the design thinking methodology has been applied to explore the ways in which Artificial Intelligence can solve various problems within the field of nutrition.

Project duration

9 weeks / 55+ screens

Tools used

Design thinking process



Offer people and health clinics an application which would help provide accurate, personalised nutritional advice with the use of AI. Once patients’ blood tests are analyised in the lab, AI will generate various diet plans, suggest aliments to include in the users’ diets, provide information about the importance of each aliment and mineral, track progress and generate timeframes.


Benefits to the customers


Benefits to the sector

Target audience research

Three main categories of nutrition content consumers:
Dietitian patients demographics:

Competitor analysis

A comprehensive competitor analysis was conducted in order to assess this project’s competitors’ strengths and weaknesses and identify a gap in the current market.

Additionally, using a competitor analysis comparison table, contenders have been analysed in terms of product quality, reputation, pricing, position on globe and user reviews.

Following the competitor analysis and considering user feedback, it was concluded the main issues are:

Personas & User stories summary

Gwen Fernsby

A lady with various food intolerances, frustrated with her diet plans.


Alexandra Loughty

A stressed and unmotivated artist struggling to maintain her weight.


Elias Dankworth

A skinny guy with no nutritional background attempting to build muscle to no avail.


Empathy mapping

User POVs

Gwen needs to find a more organised way of
viewing her dietitian’s advice and test results so that she manage her health more easily.

Alexandra needs to be able to choose her
meals so that she can eat the foods she
enjoys and stay motivated.

Elias needs a way to understand how nutrition
influences his progress so that he can improve
his workouts and physical energy.

+6 stories

How might we...

Help Gwen keep track of her test results and improve their quality?
Motivate Alexandra to maintain a
Offer Elias accessible information that
would improve his health and workouts?
+7 questions


+12 features
+10 ideas

Crazy 8s

The Crazy 8’s design method was used during this project to quickly generate ideas and attempt to find the best ways of displaying important features.

Following the first Crazy 8’s session, it was assessed that the “diet plan” frame was the most important aspect of the app, followed by the “test results” frame.

The second and third Crazy 8’s sessions focused on the before-mentioned screens and attempted to ideate different designs and features.


User journey mapping

Journey maps were created for each persona, helping clearly identify:
+ 1 iteration

Decision Matrix

User flow

+ 1 iteration

Low-fi Prototype

With the use of user flows and storyboards, a few gaps within the current design process have been noticed, such as having separate screens for “daily tasks” and “diet plan”, along with a general lack of features. In order to ensure the quality of this project, I revisited the initial steps and iterated.

This way, I was able to ideate more efficient ways of displaying information and features. This marked the point where I was confident enough to transition into the prototyping phase.

+ 6 screens

User testing #1


Six testers aged 23 to 65 with no design background were chosen to participate in think-aloud sessions in order to evaluate the prototype.

Before the study, participants were briefed on the purpose and details of the prototype, and were encouraged to be critical, with a reminder that there were no right or wrong answers.

Throughout the study, their actions and comments were observed and documented.


In terms of positive feedback, overall, participants found the prototype easy to navigate, liked the complexity of features, and described the app as “professional” and “consistent“.

However, some negative observations were noted. Participants encountered minor usability issues regarding the visibility and clarity of features.

Following these insights, a high-fidelity prototype was created to address recommendations and resolve conflicts. 

User testing #2


The same method was applied a second time in order to test the high-fidelity prototypes, build upon previous findings and address any new observations or conflicts that came up with the introduction of the high-fidelity prototypes.


Differing opinions arose on the placement of different elements, along with a further lack of clarity regarding different features.

Despite these nuances, participants found the app overall useful, logical, professional, and user-friendly.

Style guide

+ 7 iterations


To ensure a good user experience, I prioritised accessibility throughout this project. This included:

Limitations & Challenges​


This project has the potential to transform the nutrition sector by solving various problems regarding dieting, not only for customers, but also for nutritionists. The outcomes of this study show a highly positive start to the concept of this app, and a green light to proceed with its further development.

What I would do differently

What I learned

...and more!
Thank you for reading!

Diana Cristina Iliescu