Case Study · 02
Talking it out with Hearth.
A voice-first wellness app that listens to how you feel and answers with music tuned to the moment.
From a music library to a conversation.
Problem
A doctor wanted a better way to help patients in distress. In their experience, music played at certain frequencies relieves stress and helps people feel better for a while. But a distressed patient is the least able to search for the right track at the moment they need it.
Solution
Hearth replaces the library with a conversation. You say how you feel, Hearth reflects back what it heard, and a playlist tuned to your mood and heart rate is ready in one tap. The shift is from browsing for calm to being met where you are.
Impact
- 01Voice in place of a music library → a patient in distress gets calming music by saying one sentence, with nothing to search or choose.
- 02An orb that says back what it heard before playing → patients feel understood first, so they trust the music that follows.
- 03Playlists tuned to mood and heart rate → the music fits the patient's state in that moment, not a generic calm mix.
- 04One permission screen each, saying what you get, with Skip for Now → sharing voice and health data feels like a choice, not a demand.
- 05Feedback asked as a feeling, plus a session history → each session learns from the last, so care improves over time.
Designing for a moment of distress.
- Why it needed to exist. The brief came from a doctor who uses music to ease patients' stress. Certain frequencies helped patients feel calmer for a while, but there was no simple way to give them the right music when distress hit.
- What Hearth is. A mobile companion built around three promises in its tagline: Your Mind · Your Care · Your Space. One loop runs the product: talk, listen, reflect, look back.
- What shaped the design. It handles sensitive input: your voice, how you feel, your heart rate. Trust had to be designed into the first minute, not added later. Heart rate comes from the phone's health data (Apple Health or Google Fit), so it works only when the patient wears a watch or tracker. Voice also had to work in noisy or public places, which is why a keyboard option sits beside the mic.
- What was unknown at the start. We started with four open questions. Will someone in distress talk to an app, or is that too much to ask in that moment? How can Hearth sense how a patient feels without handing them a form? What would make a patient trust an app with their voice and heart rate? And how will Hearth know the music helped, and change course when it didn't?
The onboarding, in motion.
Want the full story?
This page shows the demo. The research, the directions we explored and the decisions behind each screen are better told in person. To know more about the project, let's catch up.