Hyperpigmentation-first
Dark spots, melasma, PIH, and uneven tone are the centre of analysis — not a sub-category.
Now in early access · Built for Fitzpatrick IV–VI
Stop guessing
what your skin needs.
Kira is an AI dermatologist trained on melanin-rich skin. One selfie reveals hyperpigmentation, PIH risk, melasma, and texture — then formulates a routine calibrated to your Fitzpatrick type.
The science
A skincare model that
actually looked at you.
Most skincare AI was trained on Fitzpatrick I–III faces. The result: misread ashiness as healthy, miss melasma entirely, recommend retinoids that trigger PIH on Type V–VI skin. Kira’s vision model was tuned the opposite way — concerns that disproportionately affect melanin-rich skin are the centre of analysis.
-
001
Concern-weighted analysis. Hyperpigmentation, PIH, melasma, and texture irregularities are scored independently with severity 0–10 — not bundled into a single "skin score".
-
002
Fitzpatrick-calibrated actives. Recommendation strength varies by skin type. No hydroquinone >2% for Type VI. No 10% AHA peels for sensitive PIH-prone skin.
-
003
Multi-angle synthesis. Front, left, and right scans are reconciled into one report. Dark spots that hide at oblique angles get caught.
How it works
Selfie to formulated routine.
Thirty seconds.
-
01
Three-angle scan
Front, left, right — guided in real time with face-detect. PIH and melasma often hide at oblique angles. Front-only scans miss what side angles surface.
-
02
Vision analysis
Each angle is reconciled into one report: Fitzpatrick type, skin metrics, concerns with severity, exact face zones where each lives.
-
03
Routine formulation
Morning + evening steps. Active strength tuned to your Fitzpatrick. Products picked merit-first from partner brands — never paid placement.
-
04
Weekly tracking
Re-scan weekly. Severity scores drop in numbers, not feelings. Side-by-side comparison shows your fade timeline as it actually happens.
By the numbers
The math behind the match.
30s
Scan to routine — including image upload, vision pass, and recommendation rank.
16
Distinct skin concerns surfaced individually, with severity scored 0–10 each.
3
Angles reconciled into one report. Side angles catch what front photos miss.
IV–VI
Fitzpatrick range the model was tuned for. Lighter tones also supported — just not the priority.
What’s inside
Designed around your concerns.
Fitzpatrick-calibrated
Active strength scales with skin type. No hydroquinone hammers on Type VI. No 10% glycolic peels on sensitive PIH-prone skin.
Ingredient education
Tap any ingredient. Get a three-section breakdown: what it does for melanin-rich skin, the watch-outs, and the bottom line.
Weekly progress
Re-scan weekly. Concern scores drop in numbers, not feelings. Side-by-side time-lapse shows your fade in real time.
Multi-brand catalogue
Products from curated partner brands. Ranked on concern-ingredient fit. No pay-to-rank. No house-brand bias.
In-app commerce
Cart, checkout, and order tracking inside the app. Cash on delivery across Nigeria. Card payment rolling out via Paystack.
First app that didn’t put me on retinol day one. Kira saw my PIH risk, built around it, then explained why — in language that wasn’t condescending.
Adaeze, 28 Fitzpatrick V · Lagos
For skincare brands
Scan-validated intent.
No pay-to-rank.
Kira customers don’t browse — they scan their face, identify a specific concern, and ask the AI for the right product. Your formulation reaches them through merit, measured in concern-ingredient fit and Fitzpatrick suitability.
- +15% Commission only. No listing fees, no platform tax, no upfront cost. Pay per order completed.
- =00 Equal placement. Every recommendation comes from a shuffled catalogue. Tenxix has no advantage.
- 31% Conversion benchmark. Scan-validated buyers convert ~3× better than ad traffic. We have the receipts.
READY
Your skin deserves
the right answer.
Free during early access. No subscription. Built in Lagos for skin the rest of the industry overlooked.