A library that knows the songs but not the moment.

Spotify is exceptional at telling you what you listened to. It's far less fluent in why — the late drive, the heartbreak, the first warm morning of spring. This concept reimagines the experience around emotional context: a way to find music by the feeling you're in, and to revisit the moments your music has scored.

Discovery has become a search problem.

Listeners increasingly know they want "something" but not its name. Existing surfaces optimize for retrieval — playlists, genres, recently played — which work when you already know what you're looking for. They break down at the exact moment music matters most: when you're feeling something and want a sound to match.

Make the app meet you where you are.

• Let people find music by mood and context, not just title or genre.
• Turn listening history into memory — something to return to, not scroll past.
• Reduce time-to-play without flattening discovery into the same five recommendations.

End-to-end, from research to prototype.

I led the project independently — framing the problem, running the research, and designing the full flow. I owned the interaction model, the visual exploration within Spotify's existing system, and the usability testing that pressure-tested it.

Diverge widely, then commit.

I started with listening diaries to understand when and why people reach for music, then mapped those moments to interaction patterns. From there I sketched broadly, narrowed to two directions, and prototyped the strongest before testing.

People remember feelings, not file names.

• Mood is the real query. Participants described what they wanted in emotional terms long before they named a genre.
• Music is memory. Nearly everyone had songs tied to specific people and places — and no good way to find them again.
• Choice fatigue is real. Abundant libraries created paralysis; people wanted fewer, better-matched options.

Designing for emotion without gimmicks.

I introduced a Mood entry point that blends color, motion, and language — so picking "wistful" or "wired" feels expressive rather than clinical. A Memory layer surfaces songs tied to past moments, gently, without turning private history into a feed. Every choice stayed inside Spotify's existing design language so the concept reads as evolution, not costume.

A calmer, more emotional way in.

The redesign opens with a single question — how do you want to feel? — and builds the session from there, weaving in tracks the listener already loves and the memories attached to them. Discovery becomes a mood, not a menu.

Emotion is a design constraint, not a decoration.

The hardest part wasn't visualizing mood — it was respecting it. Designing around feeling meant resisting the urge to gamify or over-quantify something deeply personal. The most successful moments were the quietest ones.

Where I'd take it next.

• Prototype an adaptive model that learns mood from listening behavior over time.
• Test the Memory layer's privacy boundaries with a wider, more diverse group.
• Explore an ambient, screen-free mode for in-car and home listening.

The kindest alarm is the one no one else hears.

Silent Wake is a concept feature for the Oura ecosystem: a haptic, light-led wake experience that rouses you during your lightest sleep phase — quietly, on your wrist, without an alarm that wakes the whole room. It's an alarm designed around two bodies, not one.

Alarms are blunt instruments.

Traditional alarms ignore both sleep science and shared beds. They fire at a fixed time regardless of sleep stage, jolting you out of deep sleep, and they're loud by design — waking partners who didn't need to be woken. For couples on different schedules, the morning is a small daily conflict.

Wake one person, gently, without waking the other.

• Use sleep-stage data to wake at the optimal moment, not a rigid time.
• Make the experience silent and personal — felt, not heard.
• Build trust: people need to believe a silent alarm will actually wake them.

Concept, interaction, and motion.

I shaped the feature from insight to interactive prototype — defining the wake logic, designing the on-wrist and in-app interactions, and prototyping the haptic-and-light choreography that makes a silent wake feel reassuring rather than uncertain.

Grounded in sleep science.

I reviewed research on sleep cycles and wake thresholds, interviewed people who share beds across different schedules, then prototyped escalating haptic patterns to find the line between "gentle" and "missable." Motion and timing were tested on-device feel, not just on-screen.

Trust is the real feature.

• People fear missing it. The biggest barrier to a silent alarm isn't comfort — it's anxiety that it won't work. The design has to earn trust.
• Escalation matters. A single buzz is easy to sleep through; a gradual build wakes reliably without shock.
• Partners want agency too. The non-waking person wanted assurance they wouldn't be pulled awake by accident.

Reassurance, designed in.

I designed a gradual haptic crescendo paired with a soft, warming light that only the wearer perceives. A pre-sleep confidence cue confirms the alarm is armed, addressing the "will it work?" fear head-on. A gentle backup escalation guarantees you'll wake — so the feature can be trusted from night one.

A morning that belongs to you alone.

Silent Wake reads your sleep stage, finds the lightest moment inside your chosen window, and brings you up through touch and light. Your partner sleeps on. The day starts on your terms, quietly.

Designing for the body is designing for trust.

Wellness features live or die on credibility. I learned that the emotional work — convincing someone a silent alarm is safe to rely on — was harder and more important than the interaction itself. Reassurance is a feature you have to design on purpose.

Where I'd take it next.

• Run an at-home sleep study to validate wake reliability across sleep types.
• Explore shared-household logic for multiple Oura wearers in one bed.
• Design a "wind-down" companion so the wake experience starts the night before.

Wellness content that doesn't hijack your nervous system.

Most "wellness" apps borrow the same attention-grabbing patterns as the platforms they claim to be an antidote to — endless feeds, streak pressure, notification noise. The Calm Edit is a content platform built on the opposite premise: fewer, better things, presented calmly, designed to end the session rather than extend it.

The wellness industry uses anxious design.

Streaks, badges, infinite libraries, and red-dot notifications create exactly the low-grade stress people downloaded the app to escape. The result is "wellness" that feels like another obligation — and content so abundant it becomes another source of overwhelm.

Design for enough, not more.

• Curate a small, intentional daily edit instead of an infinite library.
• Remove anxious patterns — no streaks, no guilt, no red dots.
• Make the experience feel like a calm ritual, with a clear, satisfying end.

Sole designer, end to end.

I owned the entire project — research, information architecture, content strategy, visual design, and the interactive Figma prototype. I defined the product principles and made the editorial and interaction calls that hold the whole experience together.

Principles first, pixels second.

I began by writing the product's anti-patterns — the things it would refuse to do — then designed within those guardrails. Concept testing focused on emotional response: did the interface make people feel calmer or busier? That question drove every iteration.

Calm is a feeling, and feelings are testable.

• Abundance overwhelms. Large libraries made people feel behind, not supported.
• Endings feel good. A defined stopping point gave people permission to be done — a rare and welcome feeling.
• Tone is everything. Soft language and unhurried pacing shifted the felt experience more than any single feature.

Restraint as the headline feature.

A single daily edit replaces the infinite scroll. Generous whitespace and slow transitions set a calmer pace. Progress is shown through gentle, private reflection rather than streaks. The whole product is designed to close — to send you back to your life feeling lighter.

A daily ritual you can actually finish.

The Calm Edit opens to one thoughtfully chosen set of content for the day, presented with space and care. When you're done, you're done — no feed pulling you back. Explore the full interactive flow below.

Subtraction is the harder design skill.

It's easy to add features; it's hard to defend an empty space. The Calm Edit taught me that the most caring thing a product can do is end well — and that designing for less requires more conviction than designing for more.

Where I'd take it next.

• Develop the editorial/curation model that decides what makes each day's edit.
• Test the "calm vs. busy" emotional metric with a larger, longitudinal group.
• Prototype a weekly reflection that stays gentle and never becomes a streak.