Forest

Market Analysis

The AI emotional wellness market is projected to grow significantly, with reports indicating a $1 billion valuation by 2025. However, existing tools primarily focus on static wellness features or clinical approaches, leaving a clear gap for advanced, emotionally intelligent AI solutions.

feel comfortable discussing personal issues with generative AI

User Research

User Survey/ In-Depth Interviews

I conducted a survey and In-Depth Interviews involving 77 participants to uncover the unique challenges faced by users with insecure attachment styles (anxious, avoidant, and disorganized) and identify how an app can support their emotional growth.

Findings

75%

feel overwhelmed by relationship conflicts but hesitate to seek professional help.

69%

avoid apps perceived as overly clinical, judgmental, or data-driven.

87%

talk to friends or family members about relationship stress.

72%

Competitive Analysis

There are many emotional support apps, but they lack a structure that keeps users motivated to continuously learn and grow.

Research Insights

Integrated Attachment Theory

Insight

Most existing apps fail to address relational and emotional growth through structured attachment-based methodologies, leaving a gap in targeted relationship-building tools.

Opportunity

By incorporating attachment theory with guided emotional and relational development, your app can uniquely position itself as a holistic solution for fostering meaningful connections and personal growth.

Gamification for Relational Skills

Insight

Gamification and habit tracking are underutilized in helping users build relational skills. Many apps focus on generic self-care or companionship without encouraging active skill development.

User Retention Challenges

Insight

Although apps with gamification and characters attract initial attention, many fail to maintain user engagement due to repetitive or shallow interactions. This results in users losing interest quickly.

Opportunity

Designing an engaging and dynamic system—where users feel challenged and rewarded for consistent participation—can significantly boost retention and long-term satisfaction.

Personas

The target users for this product are late teenagers and young adults between the ages of 16 and 34, who are navigating the

complexities of various relationships, especially in new or unfamiliar social environments.

From Static Room

to Living Forest

The immersive forest design was introduced to provide a sense of growth and connection.

Enhanced navigation by organizing features into a visually engaging and intuitive layout.

From Predefined Scenarios to Mission-Based AI Chat

Scenarios chosen upfront don’t adapt to user progress or responses during the role-play.

Visiting animals with unique traits assign personalized missions, including chat activities.

89% of users felt more engaged with character-driven missions over static activities.

From Text-Based

to Voice-Centered

72% of users felt more authentic when speaking versus selecting from predefined options.

Shifted to a voice-first chatbot with text fallback for authentic and engaging interactions.

Information architecture

Low-Fi Wireframes

Takeaways

What I learned

Recognized the value of early user testing in the design process, specifically conducting

tests immediately after creating wireframes.

Gained insight into the

importance of understanding the trends and lifestyles of target users for inclusive design.

Learned how to use MidJourney to generate necessary images, enhancing visual elements efficiently.

Captured the warmth and authenticity of human interactions, while AI technology can sometimes come across as impersonal.

Areas for Improvement or

Future Considerations

Expand the Behavioral Progress Tracker to reflect more nuanced changes in user engagement and emotional growth over time.

Refine the design of the AI-driven activities to ensure they feel natural and reinforce attachment styles effectively.

Incorporate additional feedback loops, such as visual and textual summaries, to help users understand their progress and next steps more clearly.

Intro

Therapists often find that the most common question patients ask is,

One key answer lies in having healthy relationships,

as they are fundamental to wellbeing in society.

“How can I be happy?”

By leveraging attachment theory, there’s an

opportunity to create a mental health app that offers

tailored strategies for managing stress, anxiety,

and relationship dynamics.

The Opportunity

A Fact that Illustrates the Why

Research shows that individuals with insecure

attachment styles are more prone to higher stress

levels and relationship anxiety.

Problem statement

“How might we create a mental health app that leverages attachment theory to help people develop healthier relationship patterns?”

Dynamic Gamification System

Solutions

Activity

Opportunity

Introducing gamified relational skill-building and habit tracking can encourage consistent engagement while making the learning process enjoyable and interactive for users.

User Testing & Iterations

Transform the app into an emotionally engaging and user-friendly platform by incorporating feedback and usability insights. Seven rounds of user testing with 13 participants shaped the journey to the final design.

Mid-Iteration

Final Version

Initial Version

Mid-Iteration

Final Version

Initial Version

Mid-Iteration

Final Version

Scenario-Based Role-Playing

Personalized Onboarding

Dynamic Gamification System

This app helps users improve their relationship habits through guided daily practice, gamified progress, and personalized support.

Role

Product designer

Service Provided

App Design

timeline

Nov 2024 - Feb 2025

Role

Product designer

Service Provided

App Design

timeline

Nov 2024 - Feb 2025

Intro

Therapists often find that the most common question patients ask is,

One key answer lies in having healthy relationships,

as they are fundamental to wellbeing in society.

“How can I be happy?”

The problem

Many people struggle with recurring relationship patterns but don’t have a safe space to reflect, learn, and grow.

User Research

User Survey/ In-Depth Interviews

Through user research, we discovered that 70% of users prefer discussing relationship issues with generative AI rather than seeking professional help.

Because it feels safer, more private, and less clinical.

Why?

Yet, most existing apps lack the human touch that makes users feel truly supported.

What We Can Do

Practice real-life scenarios with guided support

Build confidence in emotional conversations

Make emotional growth feel rewarding and interactive

Instead of offering just mood tracking or generic advice, we can create a space where users:

Target audience

Our target audience is women in their 20s-30s who often feel overwhelmed by relationship challenges.

If we create a mental health app that provides structured, interactive guidance through emotionally intelligent AI companions, users will feel more empowered to practice communication skills in a safe, engaging environment.

Hypothesis

Usability testing

We measured session-based progress perception.

Immediate engagement signals -> Self-Reflection Prompts at the End of the Mission & Session-Based Engagement Check.

Results (Self-Guided Format)

60%

completed all tasks

50%

felt unmotivated without structured guidance

50%

uncertain about their choices

40%

reported feeling disconnected from the experience

Iterations

Users said the first version felt generic and cold.

->Added animal characters and soft textures to create emotional warmth.

Mid-Iteration

Initial Version

Final Version

Users lacked motivation when choosing activities on their own.
→ Companion now assigns step-by-step missions, boosting engagement and emotional investment.

Mid-Iteration

Initial Version

Final Version

Users lost focus during open-ended text chats.
→ Switched to scenario-based voice chat with visual milestones to guide structured, goal-oriented practice.

Mid-Iteration

Initial Version

Final Version

Many users didn’t know their attachment style and felt overwhelmed by generic onboarding.
This helps users discover their attachment style and match with an animal guide.

They feel emotionally seen from the start, making the experience feel safe and personal.

Personalized Onboarding

Solutions

Dynamic Gamification System

Users needed long-term motivation without making the app feel clinical or boring.

Befriend

Users complete quests from their animal friends and earn rewards to decorate their forest.
Progress feels playful and meaningful, helping users build healthy habits through emotional bonds.

Scenario-Based Role-Playing

Users often struggle with real conversations and don’t know how to start.
Step-by-step practice chats help users rehearse difficult situations, with gentle guidance and feedback.
They get feedback and daily quests to apply what they’ve learned.

Activity

Unlocked activities like mindful meditation and gratitude journaling support emotional balance.
They can explore on their own pace and reinforce what they’ve practiced.

Save

Users wanted flexible tools they could use anytime, not just during quests.

Impact

Takeaways

What I learned

● Recognized the value of early user testing in the design process, specifically conducting tests immediately after creating wireframes.

● Gained insight into the importance of understanding the trends and lifestyles of target users for inclusive design.

● Learned how to use MidJourney to generate necessary images, enhancing visual elements efficiently.

● AI-powered companions provide warm, authentic interactions instead of the impersonal feeling common with typical AI technology.

Areas for Improvement

● Expand the Behavioral Progress Tracker to reflect more nuanced changes in user engagement and emotional growth over time.

● Refine the design of the AI-driven activities to ensure they feel natural and reinforce attachment styles effectively.

● Incorporate additional feedback loops, such as visual and textual summaries, to help users understand their progress and next steps more clearly.

more work

Double Agent Club

Web App

Double Agent Club

Web App

Double Agent Club

Web App

Instagram

MILA

Peace Mila

2025

Instagram

MILA

Peace Mila

2025

Instagram

Peace Mila

2025