Your face,measured.

A chatbot looking at your photo guesses. Kanon measures: 468 facial landmarks per photo, and a registry of 44 published anthropometric metrics, each compared against demographic norm tables from the measurement literature — matched to your sex, age band, and self-identified ancestry. Your report covers the metrics your capture supports and honestly flags the profile-dependent ones as pending — never padded, never faked.

The plan is built the same way: a deterministic rules engine selects interventions from an evidence-graded catalog, and every claim carries a citation. The writing is AI; the decisions are auditable rules. That is a different product from a chat answer — and the difference is easier shown than argued: see the full sample report.

Once your photos pass the capture quality checks, the analysis pipeline delivers your report in minutes. The questionnaire itself takes about seven.

The one exception is deliberate: any report that trips a safety review is held for human attention and delivered within 24 hours.

Same analysis either way — the membership is for verifying that the plan is actually working.

It is designed not to. Kanon never produces an attractiveness score or a ranking — no 1-to-10, no percentile, ever. Reports lead with your measured strengths before any opportunity, and every opportunity comes with a realistic, non-surgical action.

We also screen for signs that an appearance analysis could do more harm than good. If your intake shows those patterns, we decline your money and point you to better resources instead. That path is built into the product.

Facial geometry is biometric data, and we treat it that way. You give explicit, standalone consent before any processing — it is never bundled into the terms of service.

Raw photos and landmark data are deleted 30 days after your report is delivered unless you separately opt into progress tracking, and you can request deletion at any time. Tracking-tier data is retained only while your membership is active and is fully deleted within 30 days of cancellation or request.

No. Kanon provides educational appearance analysis — it is not medical advice, diagnosis, or treatment. Nothing in a Kanon report prescribes or doses anything.

Where the evidence points to a clinical option (prescription retinoids, orthodontics), your plan flags it as "discuss with a licensed provider" and gives you the measured findings to bring to that conversation.

We are honest about this: reports are generated by expert-calibrated AI, not read line-by-line by a person. Every report passes an automated safety gate before delivery — contraindication checks, banned-claims scanning, strengths-first ordering, citation completeness — and a random sample of delivered reports is independently audited by a human.

Any report that trips a safety signal is held for mandatory human review before you see it. What we never claim is that a person reads every single report — they don't.