What if your pet could understand you?
An AI-powered companion that listens, responds,
and creates a sense of presence through AR.
Year
2026
Role
Product designer
Scope
UX / AI / AR
Award
Spring Show 2026
01
Background
Despite constant interaction with digital products, meaningful emotional connection is still missing. As loneliness rises, users increasingly seek emotional support through technology, but current experiences fail to deliver.
02
The Real World Gap
Real pets are not accessible for many users due to time, cost, and space constraints. This creates a gap between emotional needs and realistic solutions.
Conversations feel scripted
68% of users feel AI conversations lack authenticity. Most interactions follow predictable patterns
No emotional
feedback
72% of users say AI responses feel emotionally flat systems respond with logic, not empathy.
No real presence
81% of interactions end immediately after use. No continuity or sense of shared space.
03
Research
40+ online survey forms, 10+ in-deep interviews, Ages 18–34, Focus: expectations, emotional preferences, ownership barriers
Surveys
Exploring expectations for virtual companionship
13 key questions | 40+ form received |8+ Key Insights
Julia
25 years old, International Student
“I tried emotional apps before, but they felt like a task. If I skip a day, I feel guilty instead of supported.”
Alex
27 years old, Product Manager
“I don’t want something I have to manage.
I just want something that’s there when I need it.”
10+ interviews conducted
High drop-off after initial engagement
10+ Interviews
What We Learned from Real Users
+6
Users
Jack
28 years old, Graphic Designer
I don’t want to stare at a screen all the time.
I want to feel it’s right beside me.
Eile
21 years old, Student
I want my pet to sense my emotions, not just follow commands.
04
Insights
Through the use of surveys, one-on-one interviews, and quantitative research, I ultimately identified three key insights to inform my subsequent design process.
01
Users want companionship without pressure
Connection matters more than interaction count
01
Presence matters more than frequency
Users want to feel seen, not just responded to
01
Personalization strengthens attachment
Meaning comes from adapting to each user
05
The Opportunity
These insights led to my design opportunity:
06
Introduction
So, what is PETAR?
07
How AI Works
A key part of this project was making AI feel concrete rather than abstract. So I framed PETAR’s AI interaction as a simple loop: input, interpretation, and response.
08
System Overview
Here is the system overview that introduce how the features can reflect my design decision from my insights.
09
Companion Emotional Design
Instead of treating pets as static visual assets, I designed them as emotional agents with distinct personalities, behaviors, and interaction patterns.
Each character is crafted to evoke a different emotional response. From comfort and playfulness to curiosity and calmness allowing users to form more personal and meaningful connections.
10
AI-Assisted Character Design
This page shows how I used AI as part of my visual exploration process. I began with initial sketches and then used AI-assisted exploration to test different character directions more quickly. From there, I evaluated the results based on emotional tone, silhouette clarity, consistency, and product fit. This was important because I wanted the character to feel emotionally appealing, but also readable and consistent across the product system.
11
Design System
This page shows part of the design system behind the product. I developed visual foundations, components, interaction patterns, and emotional logic to support consistency across the experience. For a project like PETAR, design system thinking was especially important because the experience spans AI, AR, conversation, memory, and character interaction. A strong system helps those pieces feel connected rather than fragmented.
5+ participants
Key findings from testing
• Emotional understanding needs to be clearer
• Interaction variety increases engagement
• Users value long-term emotional reflection
13
Iteration & Design Refinement
Based on user testing insights, I iterated on key flows to improve clarity, engagement, and emotional connection.
Result
Users were able to quickly understand the product and showed
higher initial engagement.
Result
Users could more easily identify actions and felt more engaged with the companion.
Result
Increased feature visibility and improved navigation efficiency.
14
Final Vision
Here is the final vision of PETAR as a full companion experience. This final prototype brings together the major ideas of the thesis:
emotion-aware AI, personalization, spatial presence through AR, and memory-driven continuity.
15
What I learned
Designing emotional products requires balancing technology, behavior, and human psychology. The most challenging aspect was creating interactions that felt authentic without being overwhelming. Every micro-interaction needed to serve the emotional connection, not just the functional goal.
16
Retro
Looking back, several things worked well.
Built a clear emotional product direction
Translated abstract feelings into interaction design
Created a cohesive experience across app flow and AR moments
Test with a broader range of users
Develop richer pet behavior differences in AR
Try to build this idea into a real App
Emotional design needs clarity, not just charm
Interaction variety sustains long-term engagement
Strong concepts become stronger through iteration
A companion that
doesn’t just
respond.
but understands.
PETAR explores how AI and AR can move beyond functionality
to create emotional connection in digital experiences.
Thank you for watching ：）
Check the Experience
