CONTEXT

Flourish AI

Turning Meals and Symptoms Into Personalized Gut Health Insights

As a Product Design Intern at Flourish AI, I had the opportunity to contribute to three key areas of our MVP: the meal logging process, meal logging detail page and the early exploration of third-party health kit integration. These features are central to our mission of helping users uncover dietary triggers and manage chronic gut health conditions more effectively.

In this case study, I’ll provide an in-depth exploration of my design process and prototyping work, highlighting both the “why” and the “how” behind my transformation of Flourish AI’s meal logging experience.

What is Flourish AI?

Flourish AI is a digital wellness app that helps individuals better understand and manage their gut health through data-driven insights. By combining food logging, an AI-chat bot, and third-party integrations like Apple Health, Flourish AI transforms complex biometric and lifestyle data into clear, actionable feedback. The app empowers users to identify personal food triggers, track symptom patterns, and make informed dietary and lifestyle adjustments

Who is our target user?

Role

The primary user for Flourish AI represents a wide spectrum of individuals managing digestive health, from those experiencing occasional bloating to people living with chronic conditions like IBS. These users are often health-conscious and tech-savvy, but face frustration when trying to connect the dots between meals, lifestyle factors, and symptoms.

They are not interested in calorie or macro tracking, but rather in gaining clarity about what triggers their symptoms and how to prevent them. What they want is simplicity, actionable insights, and support from a platform that can transform scattered logs of meals and symptoms into clear patterns, personalized recommendations, and everyday guidance for improving gut health.

After compiling the user interview data, I gathered 3 key insights to implement in my designs:

• Users want digestible feedback rather than reading through paragraphs of nutrition data

• Personalization drives motivation: they wanted to see trigger details tailored to their own symptoms

• Other apps did not connect their symptoms directly to what they were eatinG

Meal Detail Page

Iteration #1: Finishing the page with key information

Product Design Intern

2 Product Design Interns
1 Product Design Lead

Skills

First, I focused on improving the user journey for reviewing meals and actionable insights, two of the most high-frequency actions in the app. I started with the meal detail page (reviewing meals) as that was the most high impact, low effort to finalize for deployment.

I started by conducting user interviews with individuals managing gut health issues like IBS, acid reflux, and chronic bloating (target users). My goal was to understand:

• How users prefer to log meals (text, voice, or photo-based input)

• What information they want to see immediately after logging a meal

• How they interpret data visualizations and progress feedback

With the users’s needs in mind, I started off simple with my lo-fidelity ideas, including the meal name, some nutritional details about the meal and a description of the meal (provided by the AI in-app). Then, I went ahead and implemented a feature called the trigger score which is based on trigger ratings from our research for certain foods that can trigger IBS. This is a key visual feature that is easy for users to understand.

With this iteration I aimed to make the meal detail screen more informative and scannable. The original design already included basic meal descriptions, so I focused on surfacing nutritional and trigger-related insights in a structured way.

Using components, I created cards for each ingredient which contained any symptom insights and fixes they could employ in their specific meal via the drop down. Additionally, I color coded the trigger scores to visually indicate a “good” score against a “bad” one for quick glanceability for users. I also included the macros for each dish in the ingredient cards.

Iteration 2: Moving away from a stats heavy homescreen

Although the page was now “complete”, we felt that that opening this page as a new user would cause a lot of cognitive overload on the user. So, I went back to the drawing board to come up with some ideas on how to make this page more visually appealing while still conveying the information it needs to to the user.

In this iteration, I focused on making nutritional data more intuitive and space-efficient. Instead of showing plain numerical macros, I introduced progress visualizations to help users quickly interpret their intake relative to daily goals which especially helpful for those new to nutrition tracking.

I also simplified ingredient cards to display only the trigger score by default, with detailed insights available on expansion. This reduced cognitive load and visual clutter, allowing users to focus on what matters most at a glance.

After a round of beta testing with our first round of users, we found a key complaint:

“The visual trackers are misleading in terms of units it’s measured in since calories and protein aren’t measured at the same scale”

I added units in the circular progress trackers since the units were unclear, however, since all the units are different I opted for a different type of visualization as shown in the next screen to the right.

Here, instead of the circular progress bars, I used horizontal ones instead because it is easier to keep track of scale in that way and units as well.

I also added a “breakfast” tag for what meal this is and a way for users to also edit any details about the meal next to the name of the meal.

This was my design that was shipped!

MVP1 3RD PARTY KIT INTEGRATION

Meaningfully integrating 3rd party apps into the homescreen and app layout

Flourish AI’s goal is to make sure that all users can get actionable insights on their daily habits, not limited to food but also any data other apps collect such as apple health, cgm monitors, etc. My goal was to integrate these apps into the current layout of the app’s homescreen and display the stats we get from these apps in the most user friendly way.

I wanted to get a better understanding of how other apps approached the concept of 3rd party apps before diving into ideation for designs, so I mapped out the flows of each related apps such as Nourish, Cronometer, and Function Health to gather insights on how we could differentiate ourselves

After auditing these apps, focusing on the strengths and weaknesses of current health/nutrition apps and understanding how data drives personalized recommendations, I arrived at the following insights:

Meal Detail Logging Page

I created three user personas to determine the ideal target audience for our product. The first represents data-driven users who enjoy analyzing detailed metrics, the second captures beginners who feel overwhelmed by too much information, and the third blends both: users who track symptoms and seek to understand why they occur and how to address them. Ultimately, I focused on the third persona, designing ways to make our data more insightful and actionable rather than purely descriptive.

My first goal was so figure out where these 3rd party integration stats would even be located on our app and what they would be, relevant to gut health and IBS/GERD users. Collaborating with the research team, they said “steps”, “exercise minutes”, “sleep”, and “heart health” are most crucial to display.

Now, I just had to figure out where: the homescreen because we want users to be able to see their stats with the least amount of clicks as soon as they open the app.

The homscreen takes up a lot of empty space with no direct actionable insights immediately upon opening the app.

My goal was to remove the check-in box completely since it does not need to take up so much space and the white box below it. I also focused on displaying their main meal macros, surfaced from the meal detail page from before and also their stats from their 3rd party app of choice. This way, the user is able to view their stats immediately and click into their stats (on the insights page to read more).

I talked to my product lead and upon looking at the previous design we realized that it was too stats heavy for a user who doesn’t just want data, but also insights. Someone who may have other apps to view their macro details and activity status may want to see something that makes our app differentiate amongst others. So, I ideated on design ideas that give users more insights up front with less overload of stats.

FINAL DESIGN/PROTOTYPE

Working at Flourish AI taught me how fast-paced and deeply collaborative product design, especially in a startup environment, can be. Because I was involved in nearly every step from early research to interface design to handoff, I learned to balance iteration speed with design depth, prioritizing what would make the biggest impact within our MVP timeline.

This experience also showed me that design is rarely linear and some of my strongest ideas emerged from revisiting earlier assumptions or user pain points. At times, I’d sketch something in Figma only to realize a small design change would impact the logic of an entire flow, pushing me to think more holistically about the product!

If given more time, I’d focus on usability testing across different user conditions (chronic vs. mild symptom users) and for the meal logging process on the home screen. I would also want to focus on a journal feature for users to view not only their logged meals, but also their 3rd party integration data in one place.

Reflections

Designed and developed a restaurant website and ordering system for a Taiwanese family business, currently being deployed for launch. This project deepened my understanding of end-to-end product design. Please check it out!

Bentolicious

Iteration #2: Enhancing the page for glanceability