Why Kintsu could not use an off-the-shelf skin tool
Why it matters. The practice serves Fitzpatrick III to VI, a group most commercial skin tools serve poorly. The calibration is written for IV to VI, where those tools fail worst. A generic widget would have carried in the bias the practice sells against.
What the consultation tool does
So we built the capture ourselves: five quick frames, one after another, lit only by the phone screen.
Five frames, about six to eight seconds
A skin consultation taken on a phone, in the time it takes to hold still
The phone screen is the lamp. It changes five times while the camera watches, and the visitor does nothing but sit there.
One. DarkThe room as it is
Think of the lighting booth a dermatologist works in, shrunk down to the screen a patient is already holding.
We tried colored light first and dropped it, because it reads less accurately on darker skin, which is exactly where this tool has to be strongest.
A visitor learns which treatments fit them before they call you, and no photograph is ever kept.
- A plan first, then a tool in 48 hours. A design document and a task-by-task test plan on May 29, 2026, then results by face area.
- Running privately, July 2, 2026. Capture takes five frames in about six to eight seconds, lit only by the phone screen. Patients cannot reach it yet.
- A lid on the cost, July 4, 2026. Consultations are open to visitors in DC, Virginia, and Maryland. If that check ever fails, it lets people through rather than turning them away.
- Enquiries land in Zenoti, July 2026. Website enquiries become guest records automatically, matched on email then phone so nobody is entered twice.
- Better photos, August 20, 2026. The newer five-frame capture became the default, and an on-screen guide now shows visitors how to hold the phone.
How does a skin tool stay on the right side of diagnosis?
The tool is built so it cannot diagnose, rather than warned not to. The model may not use condition names, causes, or clinical grading, which is what keeps the tool outside the practice of medicine. It speaks in ten concerns across eleven face areas and points only at treatments Kintsu offers.
Nothing is kept. No database, no accounts, no stored photo or result. Follow one photo from the moment it is taken and it never leaves the visitor's phone, and what reaches booking carries no health detail at all.
One photo, start to finish
A skin tool with nowhere to keep a patient's photo
Three points where a photo could have been kept, and what happens at each one instead.
Step one
The visitor agrees, then takes the photo
It is taken on their own phone, and it stays on their own phone.
Step two
The photo is read once, and then it is gone
No database, no account, no stored photo and no stored result. There is nothing kept, so there is nothing to lose.
Step three
Only the treatments travel to booking
What reaches your booking page is the list of treatments that interested them, with no health detail attached.
A tool built to draw patients in cannot become the thing that leaks their faces.
The bias controls are written down. Written instructions for deeper skin tones correct a known habit of these tools. They over-flag pigmentation and under-call laxity on darker skin. Redness is left out, the least reliable signal on melanin-rich skin. Results are outlines drawn on the visitor's own photo, because regenerating an image can lighten skin tone.
Why we said no to colored-light capture
We turned it down twice, on May 31 and again on July 3, 2026. It imitated a clinical imaging booth, but a phone screen only tints the picture. It cannot separate what sits under the skin, and these models are not trained on tinted faces. It degrades worst on the deepest skin tones, where this tool has to be strongest.
What went wrong
Six weeks on the wrong capture. After the July 2, 2026 start the tool served the older single-shot capture, until our August 20 audit caught it.
A day of failed analyses. On August 19, 2026 every consultation failed because a security key never reached the build, and the release now refuses to ship without it.
A second set of eyes caught a real bug. An independent reviewer found the tool measuring head tilt backwards, before any visitor did.
What is measured, and what is not?
What we can show today. How fast it got built: written plan to a working, tested tool in about 48 hours.
Not yet measured. No business numbers exist yet, because no patient has used it. The number that will judge this tool is Zenoti bookings against the enquiries it sends. Two gaps are open. Facial hair has been read as fine lines, with a fix designed on August 20, 2026 and not yet built. No labeled image test set exists.
What happens before patients see it
Kintsu's owner cleared the tool for non-diagnosis use on September 4, 2026. Four things still stand between it and the public: attorney review of the consent wording, a Virginia Consumer Data Protection Act check, a spend cap, and medical director sign-off.