





01
Overview
MindfulMatch received 1st Place in the AI Design Challenge hosted by DubsTech Protothon 2023.
This was a 24 hour UX hackathon and was judged by top UX designers from Google, Microsoft, Adobe, and more!
Our prompt was revealed the day of, which was to design a solution that leveraged AI in solving the complex and mechanical problems faced by organizations and individuals daily.
We chose to focus on socially anxious daters and how we can improve their experience with AI.
MindfulMatch received 1st Place in the AI Design Challenge hosted by DubsTech Protothon 2023.
This was a 24 hour UX hackathon and was judged by top UX designers from Google, Microsoft, Adobe, and more!
Our prompt was revealed the day of, which was to design a solution that leveraged AI in solving the complex and mechanical problems faced by organizations and individuals daily.
We chose to focus on socially anxious daters and how we can improve their experience with AI.
MindfulMatch received 1st Place in the AI Design Challenge hosted by DubsTech Protothon 2023.
This was a 24 hour UX hackathon and was judged by top UX designers from Google, Microsoft, Adobe, and more!
Our prompt was revealed the day of, which was to design a solution that leveraged AI in solving the complex and mechanical problems faced by organizations and individuals daily.
We chose to focus on socially anxious daters and how we can improve their experience with AI.
Duration
24 Hours
Team
Product Designer - Julie Wu
Product Designer - Dana Chou
UX/UI Designer - Suki Liang
UX/UI Designer - Vivian Van
Role
Product Design
Design System
Rapid Prototyping
User Interviews
Usability Testing
02
Design Challenge
How might we design an dating app that leverages AI to help socially anxious daters in these areas?
The current dating app scene is a struggle for socially anxious individuals. Getting a date is rough! Users need to dodge scams and catfish. And once you find a good match, how do you know what to say and keep the conversation going?
How might we design an dating app that leverages AI to help socially anxious daters in these areas?
The current dating app scene is a struggle for socially anxious individuals. Getting a date is rough! Users need to dodge scams and catfish. And once you find a good match, how do you know what to say and keep the conversation going?
How might we design an dating app that leverages AI to help socially anxious daters in these areas?
The current dating app scene is a struggle for socially anxious individuals. Getting a date is rough! Users need to dodge scams and catfish. And once you find a good match, how do you know what to say and keep the conversation going?
03
Strategic Prep Before the Challenge
Before the chaos of the challenge began, I took proactive steps to ensure our team hit the ground running. Recognizing that we were a newly formed team with unfamiliar faces, I initiated a kickoff meeting to establish a strong foundation for collaboration right from the start.
By breaking the ice early on and gaining insight into each team member’s preferred working style, we were able to strategically assign roles and responsibilities.
I also started a small design system in Figma that could be updated the day of challenge. This included a customizable library for typography, color schemes, and iconography.
Organizing a structured framework ahead of time saved us a valuable hours during the challenge phase and enabled us to deliver a more cohesive UI.

Before the chaos of the challenge began, I took proactive steps to ensure our team hit the ground running. Recognizing that we were a newly formed team with unfamiliar faces, I initiated a kickoff meeting to establish a strong foundation for collaboration right from the start.
By breaking the ice early on and gaining insight into each team member’s preferred working style, we were able to strategically assign roles and responsibilities.
I also started a small design system in Figma that could be updated the day of challenge. This included a customizable library for typography, color schemes, and iconography.
Organizing a structured framework ahead of time saved us a valuable hours during the challenge phase and enabled us to deliver a more cohesive UI.

Before the chaos of the challenge began, I took proactive steps to ensure our team hit the ground running. Recognizing that we were a newly formed team with unfamiliar faces, I initiated a kickoff meeting to establish a strong foundation for collaboration right from the start.
By breaking the ice early on and gaining insight into each team member’s preferred working style, we were able to strategically assign roles and responsibilities.
I also started a small design system in Figma that could be updated the day of challenge. This included a customizable library for typography, color schemes, and iconography.
Organizing a structured framework ahead of time saved us a valuable hours during the challenge phase and enabled us to deliver a more cohesive UI.

04
Understanding the Problem
Once the challenge began and we narrowed our focus, we took some time to research and understand our users and problem space.
Considering our time constraint, we couldn't conduct user interviews so we settled for secondary research through published articles and first hand experiences from users on Reddit.
Our research revealed 3 key areas of struggle that we chose to focus our core features on:
Lack of authenticity in user profiles
Lack of communication skills
Difficulty in maintaining conversations

We also explored what existing competitors offered in this space and how we could differentiate our product.
Direct competitors like Iris, MatchMde, and AIMM used AI as a core feature in their dating app.
Secondary competitors like Tinder, Bumble, and CoffeeMeetsBagel were some of the most popular dating apps but did not have AI features (at the time of our project).
And we also researched into MeetUp, an indirect competitor for meeting people but not for the sole purposes of dating.

Once the challenge began and we narrowed our focus, we took some time to research and understand our users and problem space.
Considering our time constraint, we couldn't conduct user interviews so we settled for secondary research through published articles and first hand experiences from users on Reddit.
Our research revealed 3 key areas of struggle that we chose to focus our core features on:
Lack of authenticity in user profiles
Lack of communication skills
Difficulty in maintaining conversations

We also explored what existing competitors offered in this space and how we could differentiate our product.
Direct competitors like Iris, MatchMde, and AIMM used AI as a core feature in their dating app.
Secondary competitors like Tinder, Bumble, and CoffeeMeetsBagel were some of the most popular dating apps but did not have AI features (at the time of our project).
And we also researched into MeetUp, an indirect competitor for meeting people but not for the sole purposes of dating.

Once the challenge began and we narrowed our focus, we took some time to research and understand our users and problem space.
Considering our time constraint, we couldn't conduct user interviews so we settled for secondary research through published articles and first hand experiences from users on Reddit.
Our research revealed 3 key areas of struggle that we chose to focus our core features on:
Lack of authenticity in user profiles
Lack of communication skills
Difficulty in maintaining conversations

We also explored what existing competitors offered in this space and how we could differentiate our product.
Direct competitors like Iris, MatchMde, and AIMM used AI as a core feature in their dating app.
Secondary competitors like Tinder, Bumble, and CoffeeMeetsBagel were some of the most popular dating apps but did not have AI features (at the time of our project).
And we also researched into MeetUp, an indirect competitor for meeting people but not for the sole purposes of dating.

05
Defining the User: Raj
Meet Raj Koothrappali! He is the inspiration for our user persona.
If you're not familiar, Raj is a main character on the popular show The Big Bang Theory. He's known for being extremely socially anxious and unable to speak to women.
With the help of AI, MindfulMatch is meant to help socially anxious daters like Raj to trust and feel confident throughout the dating process.

Meet Raj Koothrappali! He is the inspiration for our user persona.
If you're not familiar, Raj is a main character on the popular show The Big Bang Theory. He's known for being extremely socially anxious and unable to speak to women.
With the help of AI, MindfulMatch is meant to help socially anxious daters like Raj to trust and feel confident throughout the dating process.

Meet Raj Koothrappali! He is the inspiration for our user persona.
If you're not familiar, Raj is a main character on the popular show The Big Bang Theory. He's known for being extremely socially anxious and unable to speak to women.
With the help of AI, MindfulMatch is meant to help socially anxious daters like Raj to trust and feel confident throughout the dating process.

06
Brainstorming Designs
With the user and problem clearly defined, our team was able to start ideating on how we might solve this issue.
We narrowed our focus to 3 key screens to start sketching for:
AI Supported Match Results
AI Supported Chat
AI Supported Video Date

With the user and problem clearly defined, our team was able to start ideating on how we might solve this issue.
We narrowed our focus to 3 key screens to start sketching for:
AI Supported Match Results
AI Supported Chat
AI Supported Video Date

With the user and problem clearly defined, our team was able to start ideating on how we might solve this issue.
We narrowed our focus to 3 key screens to start sketching for:
AI Supported Match Results
AI Supported Chat
AI Supported Video Date

07
Designing the Critical Flows
Our critical flows are designed to make Raj feel secure, supported, and confident throughout his dating journey.








Our critical flows are designed to make Raj feel secure, supported, and confident throughout his dating journey.








Our critical flows are designed to make Raj feel secure, supported, and confident throughout his dating journey.








08
Usability Testing
After our first rapid prototype, we conducted a quick usability test to check for areas of improvement within our design.
Our first design fell a little short in our goals & results:

So we improved our designs in the following areas based on initial user feedback:





After our first rapid prototype, we conducted a quick usability test to check for areas of improvement within our design.
Our first design fell a little short in our goals & results:

So we improved our designs in the following areas based on initial user feedback:





After our first rapid prototype, we conducted a quick usability test to check for areas of improvement within our design.
Our first design fell a little short in our goals & results:

So we improved our designs in the following areas based on initial user feedback:





09
Final Prototype
We built our prototype in Figma. Check it out below!
It's best viewed on desktop. Press R to restart from the beginning.
10
Conclusion
This was a really exciting and rewarding challenge for our team!
We had no expectations of winning the challenge and wanted to simply give it our all — even if that meant staying up all night to produce a quality prototype that we were proud of.
Here are some of our key insights and takeaways:
Make the most of what you have
Respecting the 24 hour time constraint of this project, we designed our app based on assumptions. However, we leveraged what we learned from articles and secondary research to support our design decisions.
Testing assumptions is key!
While we weren’t able to test with our target user, we still gained valuable insights about preferences and attitudes towards AI from testing with 2 participants. Through testing, we identified points of friction which allowed us to iterate and create a more thoughtful design!
Organize and collaborate early
Our team did a fantastic job early on to ensure the efficiency and success of our work by collaborating early and setting up our design file before the challenge even began. That just made working together during the challenge go extremely smooth.