sonder

Inspiration

We kept hearing the same thing from friends: overthinking texts, leaving hangouts feeling weirdly drained, and not really knowing which relationships were actually good for them. Paired with the FigBuild prompt about new senses, that sparked a question: what if you could literally see the emotional patterns in your relationships instead of guessing? That became sonder.

What it does

Sonder is a speculative AR wellness tool that gives you a new relationship sense, designed specifically with devices like Apple Vision Pro in mind. In a future version, it imagines combining wearable data such as heart rate variability and breathing with spatial signals from eye tracking and AR: how often your gaze meets someone else’s, how your attention lingers or avoids, how your body shifts as you move closer or farther away. In the moment, Sonder uses these inputs to sense when your nervous system is relaxing, tensing, or flattening around different people and situations. In AR, it turns those patterns into glowing constellations and aura like visuals anchored in 3D space, so you can literally see which connections tend to refuel you, which quietly drain you, and how your feelings change over time, instead of only guessing from memory.

How we built it

We built sonder primarily with Figma and Claude. We used Figma and Figma Make to design the core flows, interaction patterns, and visual language (constellation view, relationship bubbles, timelines, and AR concepts). In parallel, we used Claude as a collaborative partner to synthesize survey input, explore concept directions, and iterate on copy and interaction ideas. There’s no production code—this is a speculative UX and visual prototype focused on interaction, storytelling, and safeguards rather than implementation.

Challenges we ran into

We spent a lot of time wrestling with whether sonder is even ethical as a concept. Because it deals with emotions, relationships, and “patterns,” we had to ask: is it okay for a tool to reflect this kind of information back to someone, and where is the line between helpful and invasive? We questioned how much data about relationships should be collected at all, how to avoid sliding into surveillance or “truth scoring” other people, and whether visualizing patterns might accidentally push users to over‑interpret or obsess. A big part of the challenge was designing sonder so it centers the user’s own feelings and self‑reflection, rather than claiming objective truth about the people around them.

We also had to keep scope realistic. With just a weekend—and no actual access to sensors like HRV or eye tracking—we focused on self‑report check-ins and visual storytelling instead of trying to fake complex sensing.

Accomplishments that we’re proud of

We’re proud that sonder feels emotionally grounded, not just sci‑fi. Survey responses from other students mapped directly onto our flows—people really do struggle to know which connections nourish them and when their fear is distorting things. We’re also proud that we pushed ourselves outside our comfort zone by designing specifically for Apple Vision Pro and speculative AR. None of us had prior experience with spatial interfaces, so figuring out how sonder’s “vibe bubbles” and constellations could live in 3D space—while still feeling soft and human—was a big leap.

We’re excited about the visual language we developed: constellation views, aura‑like bubbles, and gradient timelines that communicate state at a glance without scores or heavy charts. And we’re proud that we built safeguards in from the start: consent, user control, and a clear message that sonder is a reflection tool, not a truth machine.

What we learned

We learned how much power there is in turning messy, subjective feelings into visual patterns. Even simple check-ins become insightful once you aggregate them and show them as color and motion. We also learned that speculative design lives or dies on framing: the same “sensing” idea can feel dystopian or caring depending on whose perspective you center and how you talk about it. Lastly, we saw how important softness is when designing for emotional topics—timing, microcopy, and pacing all influence whether someone would trust a tool like this.

What’s next for sonder

Next, we’d deepen research with interviews and co‑design sessions, especially with people working on self‑trust, boundaries, or social burnout. Product‑wise, we’d build a small real prototype that uses actual check‑in data to test if sonder meaningfully changes how people make relationship decisions. We’d explore more expressive check-ins (mood (+) body (+) context), refine accessibility, and continue evolving the AR layer—figuring out how ambient aura feedback could coexist with real‑world social norms and consent.