Private camera wellness interface

One check-in. A clearer personal context.

Zoetude turns a short, private camera session into an organized wellness snapshot: camera-derived observations, the context you choose, and a record you can revisit or selectively discuss with ChatGPT.

Wellness reference · Research Beta · not a medical device

Definition

What is Zoetude?

Zoetude is a private camera-first wellness interface. A 15–60 second check-in organizes on-device camera observations, your selected context, and personal trends. You can keep them on the device or approve a small, previewable Capsule for ChatGPT. It is not a medical device.

Camera rhythm

Pulse · Research Beta

A quality-gated estimate derived from subtle frame-to-frame color variation. It is a camera reference, not an ECG or clinical measurement.

Longer observation

Breathing · Research Beta

A camera-derived breathing-rate estimate can appear as the capture window grows and enough usable motion and signal context accumulates.

Beat intervals

HRV RMSSD · Research Beta

A longer capture can form a device-derived beat-interval estimate. It is not ECG HRV and is not converted into stress, readiness, or recovery scores.

Regional observation

Face Response

A raw-free, lighting-sensitive comparison of regional camera color ratios. It does not claim temperature, perfusion, inflammation, sweat, or skin condition.

Progressive protocol

What can one Zoetude check-in update?

Every completed check-in creates a usable wellness record. Longer capture windows add observations when the device has enough signal; they never change the product into a diagnostic monitor.

15 sec

Pulse and face response

The first complete window can save the check-in, camera pulse estimate, processed rhythm shape, frame context, and regional color observation.

30 sec

Breathing and segment comparison

Additional frames can stabilize the pulse estimate, compare consecutive signal windows, and add a camera breathing estimate.

45–60 sec

Beat-interval context

More usable beats can support a device-derived HRV RMSSD estimate and stronger agreement across windows. The user can still finish after the first complete window.

Read the camera method and evidence boundary →

User-approved AI context

How does Zoetude work with ChatGPT?

The connection is designed as a narrow interface, not a data dump. The user chooses the scope before anything is handed off.

Choose

Select the dates and categories that may enter the Capsule. Unwanted days or observations can be excluded.

Preview

Read the exact bounded Capsule before copying it or approving an OAuth connection. Raw camera media is never included.

Discuss

Ask ChatGPT to summarize the approved wellness context. AI output is not a diagnosis and remains the user's choice to use or ignore.

Data boundary

What stays private?

Processed on deviceRaw camera frames, face landmarks, regional pixel signals, the live waveform, beat processing, and the visual camera field.
Kept as bounded recordsCompleted check-in summaries, processed Research Beta estimates, user-selected context, local trend points, and raw-free face-response capsules.
Excluded from AIRaw video, face photos, the full 468-point face mesh, raw waveforms, frame-level rPPG samples, and unapproved dates or categories.
User controlsPreview, date selection, category selection, Google sign-in, connector disconnect, local deletion, and account-data deletion. Optional Face Memory is off by default and its bounded reference is never a photo, full mesh, or faceprint.

Built from monitoring research

Junhwan “Bradley” Kwon, PhD

Zoetude is being built by a biomedical engineer and non-contact monitoring researcher with experience across anesthesiology research, thermal imaging, patient monitoring, and camera-based physiological observation. The product keeps research estimates clearly separated from validated medical measurements.

Direct answers

Frequently asked questions

Does Zoetude upload my camera video or face photo?

No. Raw camera frames, face photos, the full face mesh, and raw waveforms are processed on the device and are not uploaded. Bounded summaries are stored or shared only according to the controls shown in the product.

Do I need a wearable or special camera?

No extra hardware is required for the camera check-in. A supported current browser and front-facing camera are enough. Imported wearable observations are optional and retain their source label.

What does Zoetude share with ChatGPT?

Only the dates and data categories you approve are placed in a bounded Capsule. You can preview it before sharing. Raw video, face images, raw waveforms, and the full landmark mesh are excluded.

Is Zoetude a medical device?

No. Zoetude is a wellness reflection interface. It does not diagnose, treat, provide emergency monitoring, or replace a validated medical device or qualified healthcare professional.

What happens if one camera estimate is unavailable?

The completed check-in still records the useful context it obtained, such as capture duration, frame observation, user-selected context, and supported channels. Zoetude does not invent a missing physiological number.

See your own private snapshot.

Start with the 15-second window and continue only when the additional context feels worth it.

Open Zoetude