Interviews with Xuefei Wang

Negotiabbles is a design studio based in the United States. We are a group of professional designers with diverse backgrounds joined by our passion for designing delightful experiences using AI. Our members come from backgrounds of hci, architecture, education, and motion design.

To me, design is the process of creating experiences that are problem-solving, delightful, and validated through user testing. The design artifacts could be a variety of media, whether physical spaces, physical, or digital products.

I enjoy designing for products that solve a need I resonate with. The experience of working on Negotiabble was memorable because it served to teach the skill of negotiation that I find meaningful, which is a core life skill I learned and practiced when I entered adulthood.

My favorite design is American architect Frank Gehry’s Prairie-style homes. He designed homes that are human-centered, inviting in nature, and each design decision is made thoughtfully and deliberately, from the material and color of the structures, down to the chairs and decorations.

I designed a digital tutor dashboard for an EdTech company, enabling math tutors to track students' learning performance, in-class behaviors, and growth trends, which helped the math tutors spend less time on manual tracking and more time on actual instruction.

My favorite tool is Figma. As a UX designer, I rely on it for everything from low-fidelity wireframes to high-fidelity interactive prototypes when mapping out the learner journey. It also makes collaboration seamless: I can share a live link with stakeholders, gather comments directly on the design, and iterate in real time instead of going back and forth over static screenshots. I really love its clean, intuitive product UX, which keeps design files organized and easy to navigate, even across complex, multi-screen projects.

I feel most creative when I'm in a room in person, brainstorming with other designers. There's something about that live, back-and-forth energy building on someone's half-formed idea, or being challenged to think differently that sparks ideas I wouldn't come up with on my own. I really believe in the power of colliding ideas: when different perspectives bump into each other in real time, it often leads to solutions that are more creative than what any one person could have designed alone.

When I'm designing, I focus most on usability and user flow first. Before I think about visuals, I want to make sure the interaction itself makes sense that a user can move through a task intuitively, without unnecessary friction or confusion. I usually start with wireframes or a rough prototype in Figma to test the flow before investing time in visual design.

Design brings up a mix of emotions for me, honestly. In the early stages, brainstorming, sketching, exploring ideas, I feel a lot of curiosity and excitement, especially when I'm collaborating with other designers and ideas start building on each other. There's also a bit of productive tension during that phase, since I'm often holding multiple possibilities in my head at once before narrowing down.

When a design I've worked on is finally realized, whether that's a prototype coming to life or a full product launch, I feel a real sense of pride and fulfillment, especially knowing that something I imagined and shaped is now actually helping people. But honestly, the strongest emotion is gratitude, for the people I collaborated with along the way, since design rarely happens in isolation. Seeing a final product reminds me how much of it came from other people's ideas, feedback, and iteration, not just my own.

A design is successful when it solves the user's problem so seamlessly that the user doesn't have to think about it. It's less about how polished something looks and more about whether it actually changes behavior or outcomes the way it's meant to and ideally, it's built in a way that can scale and evolve over time.

Usability comes first for me: can a user complete their task easily and intuitively, without confusion or unnecessary steps? I look at whether the flow makes sense, whether key actions are easy to find, and whether the user could get through the task without needing extra explanation or guesswork. If a design is visually polished but users struggle to figure out what to do, I consider that as a design without addressing the core needs, no matter how good it looks. Everything else, hierarchy, consistency, visual polish, matters, but only after I'm confident the design actually works for the user first.

Designers have a responsibility to create solutions that are not only useful, but also inclusive, ethical, and sustainable. Every design decision influences how people behave, make decisions, and interact with technology. As AI becomes more common, designers also play an important role in making complex systems transparent and trustworthy. Good design should improve people's lives while considering its long-term impact on society and the environment.

Design is becoming increasingly interdisciplinary. Today's designers work closely with AI, data, engineering, and research, making design less about creating interfaces and more about shaping intelligent systems and experiences. While new technologies continue to transform how we work, I believe human-centered thinking, curiosity, and critical judgment will remain the foundation of good design.

Our recent work has been exhibited through international design awards and exhibitions, providing opportunities to share our projects with a global audience. Looking ahead, I hope to participate in more exhibitions that encourage conversations around AI, human-centered design, and emerging technologies, bringing together designers, researchers, and the public.

Inspiration rarely comes from design alone. I enjoy learning from psychology, education, architecture, games, technology, and everyday observations. User interviews and conversations are often the greatest source of inspiration because they reveal unexpected needs and behaviors. Curiosity across different disciplines helps me discover new ways of solving familiar problems.

I would describe my design style as simple, thoughtful, and human-centered. I enjoy reducing complexity without oversimplifying the experience. My approach begins with understanding people's behaviors and goals, then exploring multiple concepts before refining them through testing and iteration. I believe the best interfaces often feel intuitive rather than attention-grabbing.

Living and studying in different cultural environments has broadened my perspective as a designer. Experiencing both Eastern and Western approaches to design, communication, and technology helps me recognize different user expectations and avoid designing for only one audience. That diversity has strengthened my ability to create products that are more inclusive and globally relevant.

I see design as a collaborative process rather than a service delivered at the end of development. I work closely with stakeholders, researchers, engineers, and product teams to understand both user needs and business objectives. Regular communication, rapid prototyping, and continuous feedback help ensure that design decisions remain aligned throughout the project.

The best partnerships begin with trust and open communication. Companies should involve designers early instead of treating design as a final polishing step. When selecting a designer, I recommend looking beyond visual aesthetics to understand how they think, solve problems, collaborate with others, and make decisions based on research and user needs.

My process usually begins with understanding the problem through research, observation, and conversation. After identifying key opportunities, I explore multiple concepts, create prototypes, and validate them through feedback and testing. Design is rarely linear, so I expect several rounds of iteration before arriving at a solution that balances user experience, technical feasibility, and business goals.

I appreciate everyday objects that quietly solve problems well. Some of my favorites include a mechanical keyboard for its thoughtful tactile experience, an iPad for sketching and reading, a well-designed desk lamp that creates a comfortable workspace, a simple notebook for organizing ideas, and a cast-iron Dutch oven that combines durability, function, and timeless design. They remind me that good design is often found in products we use every day.

Every day is a balance between research, design, collaboration, and learning. I usually spend my mornings on focused work such as exploring ideas, creating prototypes, or refining interactions. The rest of the day often involves discussions with teammates, gathering feedback, and aligning design with technical and product goals.
I also dedicate time to learning. With AI changing the design field so quickly, I enjoy exploring new tools, reading research, and looking beyond design for inspiration, from psychology to games and everyday experiences. I like ending the day by reflecting on what I learned and what questions are worth exploring next.

Focus on understanding people before designing for them. Trends and tools change quickly, but curiosity, empathy, and critical thinking remain valuable throughout a designer's career.
I also encourage young designers to learn beyond design. Fields like psychology, education, and technology often inspire more original ideas than studying interfaces alone. Finally, embrace AI as a creative partner, but never let it replace your own judgment. The best designers are defined by the questions they ask, not just the tools they use.

The most rewarding part of design is creating experiences that make people's lives a little easier or more meaningful. Seeing an idea become a real product is always exciting.
The challenge is that design rarely has a single correct answer. Every project involves balancing user needs, business goals, technical constraints, and limited time. While that uncertainty can be demanding, it is also what makes design an endlessly interesting profession where there is always something new to learn.

Always design for people, not for features. Every design decision should solve a real problem or improve someone's experience. If a feature adds complexity without creating meaningful value, it is usually better left out. Simplicity is not about removing functionality—it is about making the right functionality feel natural.

Curiosity is the most important skill because it drives continuous learning. Strong communication is equally essential, since designers need to collaborate across different disciplines and explain the reasoning behind their decisions. Finally, empathy and critical thinking help designers understand people's needs and translate them into practical solutions.

My primary design tools include Figma for interface design and prototyping, along with AI tools that help with research, brainstorming, and rapid iteration. Beyond software, I draw inspiration from psychology, behavioral science, games, architecture, and everyday observations. Conversations with users and teammates are often just as valuable as any digital tool.

I usually divide projects into small milestones instead of focusing only on the final outcome. Early stages are dedicated to exploration, while later stages emphasize refinement and validation. I also reserve time for feedback throughout the process, because identifying issues early is much more efficient than making large revisions at the end.

There is no fixed timeline because every project is different. A simple concept may take only a few weeks, while more complex products involving research, testing, and multiple iterations can take several months. Rather than aiming for speed alone, I believe successful design comes from continuous learning and refinement throughout the process.

One of the questions I hear most often is how we balance creativity with practical constraints. My answer is that good design is rarely about choosing one over the other. The most successful solutions are those that satisfy user needs while remaining feasible, sustainable, and aligned with business goals.

One of the most valuable experiences has been working on interdisciplinary projects where designers, engineers, researchers, and stakeholders collaborate from the beginning. Those experiences reinforced that great products are built through continuous communication and shared problem-solving rather than isolated design decisions.

Our work spans startups, technology companies, educational initiatives, and research-driven projects. While some collaborations remain confidential, they all share a common goal of creating thoughtful, human-centered digital experiences.

I enjoy designing products that combine emerging technology with meaningful human impact. Projects involving AI, education, healthcare, or interactive learning are especially rewarding because they present complex challenges while offering opportunities to create lasting value for users.

We hope to continue exploring how AI can support more intuitive, inclusive, and engaging digital experiences. Alongside new product collaborations, we also look forward to expanding our research and experimenting with new interaction paradigms that help people learn, communicate, and make better decisions.

Most of our projects are collaborative. We believe the best ideas emerge when designers, researchers, engineers, and domain experts contribute their different perspectives. Collaboration not only strengthens the final outcome but also encourages continuous learning throughout the design process.

We are currently exploring several concepts related to AI-assisted interaction, personalized user experiences, and intelligent design systems. While it is too early to share details, these projects continue our interest in making advanced technology more approachable, transparent, and human-centered.

The easiest way to connect with us is through our LinkedIn. We are always happy to discuss design, research, collaborations, and opportunities to create meaningful products together.

Design is ultimately about understanding people. Technologies, platforms, and trends will continue to change, but curiosity, empathy, and a willingness to keep learning remain at the heart of every successful project. We hope our work encourages conversations about how thoughtful design can create more positive and accessible experiences for everyone.

Negotiabbles addresses the gap by replacing passive learning with active, voice-driven practice. Most Social and Emotional Learning (SEL) tools teach children about emotions through videos or multiple-choice scenarios. Negotiabbles puts children inside a real conversation where their words have immediate, visible consequences, combining AI role-play, real-time emotion visualization, and fairytale distancing, in a way no existing tool does. What truly differentiates us is the emotion visualization system. Rather than labeling feelings with text or static icons, we built a living visual language where warm colors and angular shapes signal tension, while cool tones and rounded forms communicate calm. Children absorb it intuitively through play, without instruction.This was confirmed powerfully during usability testing. The moment that surprised us most was watching children spontaneously narrate the emotion bubbles mid-conversation, pointing at the screen and saying "see, he's getting angry now," accurately reading emotional states through color and shape alone, before we had explained the system at all. We expected to guide children through the visual language. Instead, they taught themselves. It revealed something that has since shaped our broader thinking: Children engage far more deeply when emotions have a visible and concrete form rather than remaining abstract beyond their reach.

The voice-first decision was both pedagogical and deeply human. Negotiation is a spoken, real-time skill, and teaching it through taps or text would be like teaching someone to swim without water. When a child speaks aloud to advocate for themselves, they practice the real cognitive and emotional experience of finding words under pressure, regulating tone, and responding to another voice, which is exactly what they need in a playground dispute or classroom disagreement.One of the design challenges we faced was handling silence and hesitation. When a child freezes mid-conversation, that moment carries real emotional weight. In peer interactions, silence can feel like failure, and we couldn't let the app replicate that. Our solution was to deploy AI as a sensitive, in-the-moment guide. When hesitation is detected, the AI character offers a gentle contextual nudge like "how about asking why they feel that way?", keeping the conversation alive without breaking immersion or undermining the child's sense of agency.The feedback loop is deliberately multi-sensory: the AI responds verbally, emotion bubbles shift in real time, and on-screen captions support younger readers, delivering emotional, auditory, and visual feedback simultaneously. Across the K–8 range, the system scales with the child. Younger learners receive shorter exchanges and more frequent prompts, while older children navigate complex scenarios with greater nuance and less scaffolding.

One of our core design principles was that children should experience Negotiabbles as a story and game first, not as a lesson. While the project draws heavily from developmental psychology, negotiation theory, and AI-powered personalization, we were careful not to expose that complexity directly to young users. Instead, we translated research insights into playful interactions, memorable characters, and emotionally engaging scenarios.This was especially important because negotiation often involves disagreement, conflicting perspectives, and moments of tension. For many children—and even adults—these situations can feel intimidating. We wanted to create an environment where children could safely explore difficult conversations, make mistakes, and experiment with different approaches without fear of judgment or failure. The goal was not to teach them the “correct” answer, but to help them build confidence in navigating differences.The AI works mostly behind the scenes, adapting conversations and feedback to each child's responses, while the interface focuses on curiosity, imagination, and positive reinforcement. We often described our approach as “hiding the vegetables in the soup.” The science provides the structure, but the child experiences a safe, playful world where learning happens naturally through exploration. We believe children are more willing to engage with challenging social situations when they feel emotionally secure, and that balance between rigor and warmth became a guiding principle throughout Negotiabbles.

The most difficult design decision was determining how much freedom to give the AI.Negotiation is inherently unpredictable. To help children develop real-world skills, conversations needed to feel authentic, dynamic, and responsive rather than following scripted decision trees. Yet fully open-ended AI interactions risked overwhelming children, drifting away from learning goals, or creating inconsistent experiences.This challenge led us to rethink AI not as a conversational partner, but as an adaptive learning facilitator. We designed AI-powered characters that respond uniquely to each child while remaining grounded in carefully structured negotiation frameworks and age-appropriate story scenarios.Working through this tension fundamentally shaped the product. We realized that children learn negotiation best when they are given genuine agency within safe boundaries. The result is an experience where every conversation feels personal and immersive, while still guiding children toward empathy, reasoning, compromise, and self-advocacy. It also transformed our team's understanding of educational AI: effective learning does not come from giving children the right answer, but from creating the right space to discover it themselves.

One of the most challenging questions we faced was how to make emotions visible without reducing them to labels, scores, or emojis. Early explorations relied on familiar approaches: emotion icons, mood meters, and text-based feedback. While children could recognize these signals, they struggled to understand how emotions influenced the flow of a conversation. We realized that negotiation is not a static emotional state. It is a dynamic relationship between people. This insight led us to the bubble system.We wanted to create a visual language that could be understood intuitively across ages, cultures, and reading abilities. Drawing from color psychology, emotional design research, and social-emotional learning frameworks, we began mapping emotional qualities to visual properties. Saturated colors communicate stronger emotional intensity, while softer tones suggest calmness and openness. Rounded forms feel approachable and cooperative, whereas sharper contours express tension, assertiveness, or conflict. Motion became equally important: faster movements indicate emotional activation, while slower, smoother behaviors communicate stability and reflection. Transparency helps reveal confidence, uncertainty, and emotional clarity.The hardest part was balancing expressiveness with simplicity. Early versions were either too abstract for children to interpret or too literal to feel engaging. Through multiple rounds of refinement, we continuously adjusted color ranges, shape transformations, animation timing, and interaction behaviors. We also paired bubble states with character animations and narrative context to reinforce emotional meaning from multiple channels.Over time, the bubbles evolved from a visual effect into the core interaction language of Negotiabbles. Rather than telling children how they feel, the system allows them to see emotions unfold, collide, soften, and grow throughout a conversation. The result is a living emotional mirror that transforms invisible social dynamics into something tangible, playful, and learnable.

Negotiabbles has a multiplayer gameplay mode that simulates low-stakes practice with real people, where two or three children can practice negotiation with each other in realistic scenarios and see their conversations visualized on screen. The AI serves as a mediator that prompts questions toward a resolution. This gameplay mode makes Negotiabbles playable across a wider range of environments — at home with siblings, in the classroom with classmates, and even during real-life conflicts, with AI helping to guide resolution in the moment.We plan to expand Negotiabbles’ educational scaffolding by personalizing the multiplayer learning progression according to each player’s age, personality, learning goals, and the specific scenarios they encounter day to day. A more granular progression could sequence concepts with both players’ specific needs in mind, scaling complexity accordingly and keeping the focus on each player’s individual needs and personality.In the future, we also envision Negotiabbles deepening its AI mediation layer as the platform matures. The AI mediator would no longer be a passive participant, but an active pedagogical presence — explaining its reasoning and narrating its own interventions, teaching children to recognize negotiation patterns in real time and preparing them to navigate real-world conflicts.In classrooms, Negotiabbles could provide a teacher-facing dashboard that highlights children’s learning progress and surfaces anonymized, class-level insights about which negotiation strategies children struggle with most and which conflicts arise most frequently. School counselors could use longitudinal negotiation profiles to identify students needing deeper social-emotional support, embedding the platform within a school’s broader wellbeing infrastructure.

The Negotiabbles team’s guiding principle around data and privacy is that data should serve the child first, transparency should be age-appropriate and contextually meaningful, and the architecture should be designed so that what is measured cannot easily be repurposed beyond its original pedagogical purposes.With the app’s pedagogical goals in mind, we aim to surface behavioral and performative data that is informative for children and their caregivers, and keep personal data and raw logs private.The team made a deliberate choice to give children direct, age-appropriate access to their own data in visual form — for example, the post-session summary and children’s negotiation personality profiles. This treats children as active participants in their own development rather than passive subjects of measurement. Parent-facing summaries could extend this transparency to caregivers without exposing raw behavioral logs, translating data into actionable developmental language instead.Personal data and raw logs — such as granular emotional data captured during peer-to-peer multiplayer sessions — should be kept private from players and caregivers alike, as they could be misused if taken out of context. A child’s pattern of emotional escalation in conflict situations is developmentally useful information for the app’s scaffolding engine, but surfacing it to parents or teachers without careful framing risks labeling or stigmatizing children based on in-app behavior that may not reflect their real-world character.

One of our earliest decisions was that the growth-tracking system should never feel like a report card. Through our research, we found that negotiation is not a skill that improves along a single linear path. Different situations require different approaches, and effective negotiators develop a range of strategies rather than a single “best” one. A child who practices compromise in one conversation and self-advocacy in another has not become more or less successful—they have expanded their toolkit.This perspective shaped how we visualized growth within Negotiabbles. Instead of focusing on scores, rankings, or performance metrics, we designed the profile to highlight exploration and diversity of strategies. Children can see which approaches they have practiced over time, such as empathy, reasoning, compromise, or self-advocacy. Growth is represented as an expanding landscape of experiences rather than a number increasing. The visual language also builds upon the Emotional Periodic Table, using colors, shapes, and motion to help children recognize patterns in their communication style without attaching value judgments to them.During testing, we found that children were naturally curious about these visual profiles. Rather than asking whether they had achieved a higher score, they were more interested in discovering new strategies and seeing how their profiles evolved over time. Caregivers responded positively as well, particularly when growth was framed as reflection rather than evaluation. Early concepts that relied on ratings and scores felt overly academic and sometimes created unnecessary pressure. Narrative summaries and visual storytelling proved far more effective, helping caregivers understand a child's development while preserving a sense of curiosity, confidence, and enjoyment.Ultimately, our goal was to encourage children to view negotiation as a lifelong practice of learning, adapting, and understanding different perspectives. By focusing on growth, exploration, and self-awareness rather than performance, we were able to create a system that motivates continued engagement without introducing anxiety or fear of failure.

One of the most interesting challenges was designing a system that could simultaneously support learning for children and content creation for the team. The Emotional Periodic Table was originally created to help visualize the emotional states, interpersonal dynamics, and negotiation tensions that sit at the heart of Negotiabbles. We wanted to make these normally invisible aspects of communication tangible so children could better understand both their own feelings and the impact of their actions on others.As the project evolved, we realized that the same framework could also serve as a shared language for designers, writers, developers, and AI engineers. When creating new scenarios, the team needed a consistent way to represent emotions, intentions, and interaction patterns. The table helped define how different emotional qualities could be expressed through color, shape, and motion, ensuring that new content remained coherent while still allowing for creativity and variation.What made this possible was that the system was built on perceptual patterns that children already intuitively understand. Warm, saturated colors often feel more intense, while cooler tones feel calmer. Rounded forms tend to feel approachable and safe, while sharper forms can suggest tension or disagreement. Rather than teaching children an abstract code, we built upon visual instincts that already exist across many everyday experiences.The balance between rigor and simplicity came from keeping the structure consistent while allowing the experience to remain playful. Children never need to study the framework directly. Instead, they encounter the same visual relationships repeatedly through stories, characters, and conversations. Over time, these patterns become familiar and meaningful without requiring formal instruction. In that sense, the Emotional Periodic Table functions much like the grammar of a language: it provides structure behind the scenes, while the learner engages naturally through use.By making emotions and negotiation dynamics visible, the framework supports both scalable content creation and meaningful learning. At the same time, transforming these abstract social concepts into colorful, interactive experiences helps make negotiation feel more approachable, engaging, and enjoyable for young learners.