INTERN PROJECTCreate and chat with diverse avatars for emotional support.
Conch. AI is a product based on Large Language Models (LLM), designed to allow users to create digital avatars for companionship and emotional connection.
My contributions to this project include conducting extensive domain research to understand the current interactions in AI chatbot apps, crafting user personas to understand our target users' needs for emotional support from AI, and designing interactions that enable target users to easily create their own characters and engage in immersive chats with emotional feedbacks.
Our team has completed most of the design and received positive feedback from users. The product successfully attracted 34,000 initial users and secured $5 million in funding.
Timeframe: Sept. 2023 -Present
My Role: UX Researcher, UX Designer
Team: 3 Designers, 2 Researchers
Tool: Figma, Photoshop
Mentor: June Chang
Beta app needs Refinement
Conch AI allows users to generate personalized virtual characters and earn income when other users use the avatars they have created. The initial version has completed its closed beta testing, with the beta version being relatively basic. Before the official public release, there are still many features that need further refinement.
Overview
USER’S VOICE
“The on-boarding process of creation is not friendly. The draft will not be saved after accident drop out.”
“The customization of avatar is limited. I wish to have more choices.”
“I’m looking for AI chat that can dominantly interacte with me.”
“I feel confused when calling and trying to talk with them.”
THE SOLUTION
Discovery
Kick off
Our design lead did not give us a specific design scope at the beginning stage of the project. Without knowing the scope and the scenario we would be focusing on. We explored 3 potential user groups in order to learn the importance of AI as companion and how it could aid them in daily lives.
Teenager
Their cognitive and emotional capabilities are still developing.
Middle aged
They are often busy with the dual demands of work and family.
Retired Senior
Newly retired seniors seek companionship and embrace technology for support.
After conducting semi-structured research with 7 users, we discovered that many newly retired seniors are experiencing a sense of void and loneliness. As a response, we decided to include Gen Z, who are around the age of their children. The target user groups now include young adults with depressive tendencies and elderly individuals who often feel lonely.
OUR USERS
User Requirements
Additionally, insights from the initial user testing highlighted users' primary requirements:
Offering companionship and emotional support, with a specific focus on teenagers experiencing depression and lonely seniors.
Assisting users in creating and sharing personalized virtual characters.
Facilitating interaction and socialization between users and AI bots.
After narrowing down the target group of users, we tried to understand what troubles our users may encounter and in what way they want to get emotional support from AI. Here is the story of Alex with our app.
USER JOURNEY
Through this process, we found that it's evident that users encounter issues with character creation and interactions, which are their primary areas of interest. These will be the main focus in our design process. Addressing the drop-out rate is also crucial for an improved overall user experience.
Before brainstorming for innovative solutions, I conducted an in-depth analysis of 6 related AI chatbot apps to discover our opportunities and understand users' needs.
Summarizing from this round of competitor analysis, I learned that those products all aim to connect humans and AI technology, but some lack an intuitive experience and user engagement. Therefore, this is what we tried to improve in our later design.
COMPETITIVE ANALYSIS
How might we
enable users to effortlessly create an avatar and foster emotional engagement with them?
Ideation
I worked closely with our group and conducted review meeting with developers every week to make sure that our ideation and the direction of our brianstorming is on track and technologically applicable.
In the process of updating user flow, we focus on intuitive avatar creation and expressive emotions. As this is an actual product, prioritizing its Return on Investment (ROI) is crucial. Our primary focus is to stimulate user engagement, foster content creation, and drive in-app payments to generate revenue.
UPDATING USER FLOW
Iteration
In this phase, my focus is on intuitive creation and engaging chat with avatars. Collaborating with developers, we craft prototypes for user testing, emphasizing technical accessibility and user-friendly design.
Based on feedback and reviews, we iterate on the user flow through 3 rounds and improve visual style and motion effects in 2 rounds.
Wireframes
I tried multiple variations of our primary features to explore possibilities. I mainly experimented with interaction patterns and user interfaces that work best for users with ease.
Testing 1.0
“
How to quickly start tryouts?
The process seems tiring.
Design B got a majority of votes! Users find it easy to understand and initiate the creation process with basic steps condensed onto the same page.
Testing 2.0
“
What is MBTI?
How to get a explanation?
Design A got a majority of votes! According to users, they don't need to pop up the creation process and then come back, making the process more coherent. Also, tabs on the left and right are more user-friendly than a progress bar.
“
What are the conch coins?
Where can I get them ?
Design B got a majority of votes! Users can directly access the purchase page while chatting and receive a notification when they run out of coins. This process promotes in-app purchases and enhances community awareness.
Testing 3.0
“
It looks like a specific character.
The color changing in the back is burdensome for developing.
Design A got a majority of votes! Considering development feasibility, I opted to simplify the background page and emotion effects, focusing solely on precise eye contact. This approach eliminates ambiguity between different avatars.
Make sure to pass contrast check.
When designing for this specific group, seniors, it's crucial to consider the decline in eyesight and cognitive abilities. To address their needs effectively, we've prioritized a design that is both readable and easy to interact with.
ACCESSIBILITY DESIGN
BEFORE
AFTER
To evoke the concept of magic conch, creating a sense of users landing on a mysterious bay, we incorporated elements such as bubbles and curvy lines to convey the oceanic vibe.
MOTION DESIGN
Pop-ups
The bubbles splash and disappear, delivering a delightful surprise to users when they achieve milestones.
Calling emotions
These motions effects enables users to interact with avatars and receive immediate emotional feedback.
Thinking
Connecting
Dizzy
Happy
Sleepy
Wandering
Scared
Suprised
Speaking
Angry
Sad
HI-FI COMPARISON
Final Design
ExploreLanding on conch bay, where you can explore characters and create memories!
Users can preview our community before registering, and upon joining, they will be guided to explore characters.
CreateLet’s create a personalized avatar! Don’t forget to set a price to get revenue.
Users have the flexibility to set identity, personality, introductions, and upload voice for their avatars. Additionally, we encourage advanced users to further customize through coding options.
ChatBeyond just chatting!
The AI-generated response guides users seamlessly on their first try. Users also receive immediate emotional feedback, fostering a connection with the avatar and enhancing intimacy.
ProfileManage you characters and revenue.
Our community encourage creation from everyone, We hope users actively engage and share their creations with others.
Reflection
Unveiling the Power of Collaboration
This is my first time experiencing the entire life cycle of an actual product. Throughout the process, I closely collaborated with our product manager, developers, and other designers. This unique experience has taught me skills and lessons beyond design. Witnessing the transformation from concept to a tangible product highlighted the power of effective teamwork. The developers' dedication to refining the app's functionality, coupled with our focus on user experience, resulted in a product that exceeded expectations.
My favorite part about this project was our focus on seniors and lonely young adults. Unlike other AI companion apps focused on romantic emotional support, our emphasis on real-world social challenges brought a sense of purpose to what we were doing. Moreover, this product has reassured the value of simplicity. Contrary to the belief that more choices enhance the user experience, I've realized that it can often overwhelm and tire users. An intuitive and user-friendly experience, I found, is one that doesn't burden users with too many decisions.
Overall, this project wasn't just about technology; it was a reminder of the importance of keeping things simple and meaningful, particularly when dealing with the intersection of technology and human emotions.