Step 01 · Capture
Point. Snap. Done.
Mark the area on a body-map. A guide frames each shot.
Cutis · Clinical AIAccuracy
Accuracy
for every hue.
Smartphone dermatology screening, calibrated across all six Fitzpatrick tones, and verified on every release.
Status
Private beta · Mid-2026
Coverage
Fitzpatrick I–VI
01 · The gapWhy this exists
A field designed for some skin, not all of it.
Most dermatology AI is trained on light skin. Public benchmarks rarely hold more than a handful of Fitzpatrick V and VI cases per condition, so models that look strong on lab metrics quietly underperform on darker tones in practice. The communities those models fail are often the same ones with the longest waits, the fewest specialists trained to read their skin, and the worst outcomes when something is missed. A tool meant to widen access can narrow it instead.
3billion
People worldwide without access to a dermatologist.
Access
35days
Average US wait for a routine dermatology visit. Over six months in much of Canada.
Wait time
VI
Where dermatology AI fails most, and where access gaps are widest.
Equity
02 · MethodHow Cutis works
Three guided photos. One calm answer.
Step 02 · Analyze
Results in seconds.
Urgency, confidence, and guidance, calibrated for your tone.
Step 03 · Connect
A real doctor, if needed.
Photos attached. Board-certified dermatologist on the other end.
03 · FairnessThe promise
Built for every hue
Fair on every tone. Verified every release.
We measure accuracy per Fitzpatrick tone, every release. Models that regress on any tone do not ship.
Malignancy recall
97.8%
Held-out, phone-realistic test set.
Recall · Fitzpatrick V & VI
1.000
Where dermatology AI usually fails worst.
04 · StandardsTrust & safety
Built like medical infrastructure, not a demo.
HIPAA-aware
Encryption at rest and in transit. Business Associate Agreements with our cloud provider.
Granular consent
Three-bucket consent for data use. Revocable any time, no questions asked.
Human in the loop
We screen and triage. Diagnosis always comes from a board-certified dermatologist.
Auditable by design
Every prediction is logged with its model version. Reproducible. Reviewable.
05 · TeamEngineers behind Cutis
Clinical AI, shipped before.
Built by engineers from Amazon and Square.
Teams that ship payments at billions-per-day scale and eleven-nines storage durability, now applied to healthcare that has to be right the first time.
ex-Amazon ex-Square Clinical AI · 100k+ patients HIPAA · FDA workflow
06 · DetailCommon questions
Be among the first.
Early access opens in waves. Join the shortlist; we'll write when it's your turn.
No spam · One email · That is it