Clarity Engine
How the engine works.
Most supplement quizzes are a funnel with questions painted on it. Ours is a scoring system you are allowed to inspect: deterministic rules choose the products, language only explains the choice, and the whole thing is built to be able to say no.
Rules decide. Language explains. Every recommendation traces back to a rule you could read.
The path from answer to recommendation
Six stages, one direction.
Nothing is skipped and nothing runs out of order. If a stage produces no result, the stages after it produce nothing rather than guessing.
-
01
Signals in
The assessment asks about goals and lifestyle — how you sleep, how you train, how you eat, what you already take. It never asks what condition you have. Every answer becomes a scored signal against a fixed vocabulary of six goal domains and twenty-six specific outcomes.
The vocabulary is append-only: keys are never renamed or repurposed, so a routine generated last quarter still means the same thing today.
-
02
Safety gate
Before any product is scored, one question checks for circumstances where a supplement recommendation is not ours to make. If it triggers, the pipeline stops here and returns education instead of products.
Hard stop in the code path, evaluated before the scoring matrix runs.
-
03
Deterministic scoring
Each product carries weights against the same vocabulary. Your signals are multiplied by those weights and summed. A domain score is its own signal plus every specific outcome underneath it, so answering only a broad goal still finds precisely-tagged products, while a specific answer takes precedence when you give one.
Pure functions, unit-testable, reproducible. The same answers always produce the same list.
-
04
Two lists, not one
Above the threshold, a product enters the recommended list with the reason it matched. Where your answers point against a product — stimulants for someone already high on caffeine, a duplicate of something you told us you take — it enters the not for you list with that reason stated. Anything neither strongly for nor against simply does not appear.
Recommendations are capped deliberately. A longer list is a worse answer, not a better one.
-
05
Language layer
Only now does a language model see anything, and what it sees is the finished conclusion plus an anonymous profile — never your personal details. Its job is to write the reasoning in plain English. It cannot add a product, remove one, or change a dose; those decisions were already made and are passed to it as fixed input.
If the language layer is unavailable, the report still renders from templates. The recommendation itself never depends on it.
-
06
Compliance scan
The written output is scanned before you see it. Any language that drifts toward treating, curing, or preventing disease causes the generated text to be discarded and a reviewed template used instead. Structure and function only, every time.
Second-pass scan on generated text; failure mode is fallback, never publication.
Boundaries
What the engine will not do.
These are constraints in the system, not promises in the marketing. They are the reason the recommendations can be trusted at all.
It will not diagnose.
The assessment asks about goals and habits, never about conditions or symptoms as a medical history. It cannot tell you what is wrong, and it does not try.
AI never picks the products.
Product selection and dosing are decided by rules that can be printed, reviewed, and explained to a regulator. The language model writes; it does not choose.
It will not recommend past a safety gate.
If your circumstances call for a professional, the engine says so and stops. There is no path in the system that produces a product list after a gate trips.
It will not reason about an ingredient it has no record for.
An ingredient becomes available to the engine only once it has an approved fact record — dose range, evidence level, interactions, cautions. No record, no recommendation.
Safety gates
When the engine stops.
One question near the end of the assessment can end it. Selecting any of the following stops product scoring entirely.
- I'm pregnant or nursing
- I take prescription medication daily
- I'm under 18
- I'm managing a diagnosed medical condition
You receive general educational information and a clear recommendation to speak with a qualified healthcare professional before starting any supplement. You do not receive a product list, a dose, or a routine. This is a hard rule in the code path — it is not a preference the language layer can talk its way around.
Two depths
The short assessment and the full one.
Both run the same scoring engine. The difference is how much you tell it and how much of the reasoning you get back.
| Find Your Routine | Full assessment | |
|---|---|---|
| Questions | 6–8, about two minutes | 20–25, including what you already take |
| Engine | Rules, scored in your browser | Same rules, plus a written report |
| Output | Goal profile, a short recommended list, and what you can skip | AM/PM routine, the reason for each item, the full not-for-you list, and links to the certificates |
| Your data | Anonymous, kept in your browser | Saved to your account so you can re-test and compare |
Your answers
What is stored, and where.
Health-related answers are sensitive data. They are handled under our Privacy Policy, and you can have them deleted.
- The short assessment
Answers stay in your browser. Nothing is written to an account unless you choose to save them.
- The full assessment
Answers and the resulting profile are stored against your customer account so a later re-test can be compared with the first. Consent is explicit, and you can withdraw it.
- The language layer
Receives the finished conclusion and an anonymous profile. It does not receive your name, email, or order history. Processing is carried out on our behalf by our model and infrastructure providers, as described in the Privacy Policy.
- Deletion
Write to info@wgoto.com and we will delete your assessment data. [Standard response time to be confirmed.]
Limitations
What this is, and what it is not.
Read this section as carefully as the rest of the page. It governs how the assessment and its output may be relied upon.
-
Educational purpose only
The Clarity Engine, the assessments, any report or score it produces, and this page are provided for general educational and informational purposes about dietary ingredients. They are not medical advice, diagnosis, or treatment, and they do not create a doctor–patient or other professional relationship between you and WQ.
-
Consult a professional
Always speak with a qualified healthcare professional before starting, stopping, or changing any supplement, particularly if you are pregnant or nursing, under 18, taking prescription or over-the-counter medication, managing a diagnosed condition, or scheduled for surgery. Never delay or disregard professional medical advice because of something you read here. If you experience an adverse reaction, stop use and seek medical attention.
-
No guarantee of results
Individual responses to dietary ingredients vary with genetics, diet, activity, sleep, medication, and other factors outside the assessment. A recommendation is a starting point built from the information you provided; it is not a prediction, a promise of any outcome, or a substitute for your own judgment. WQ makes no representation that any result described in published research will occur for you.
-
Accuracy and completeness of your input
The output can only be as good as what it was told. Incomplete, inaccurate, or out-of-date answers — including omitting a medication or a condition — will produce recommendations that do not fit you. You are responsible for the accuracy of the information you provide and for reviewing any routine before you act on it.
-
Research citations
Where we cite studies, we cite research on an ingredient, at a stated dose and form, in a stated population. Citing it is not a claim that a WQ product was the subject of that study, that its dose or form matches, or that any finding applies to you. Evidence levels shown in our ingredient records describe the state of published research, not the expected effect of a product.
-
Automated processing and its limits
Product selection is performed by deterministic rules; a language model is used only to phrase the explanation, and its output is scanned before display. Generated language can still contain errors. Where a written explanation and the underlying rules disagree, the rules — and the product label — govern. Nothing here constitutes automated decision-making producing legal or similarly significant effects.
-
Allergens, interactions, and the label
Always read the product label before use. Labels carry the current ingredient list, allergen statement, serving size, and cautions, and the label governs in the event of any difference from this site. Supplements can interact with medication and with each other; the engine flags known interaction cautions it holds a record for, and its silence is not an assurance that no interaction exists.
-
Availability, changes, and jurisdiction
The engine, its scoring, and this page may change as ingredient records and published evidence are updated; an earlier report is a record of that moment, not a standing commitment. Products, prices, and availability are subject to change, and products are offered only where we are able to ship. This page is written for visitors in [jurisdiction — to be confirmed] and is governed by the terms in our Terms of Service. [Governing law and dispute-resolution terms to be confirmed with counsel.]
-
Limitation of liability
To the fullest extent permitted by applicable law, WQ and World Q Corporation are not liable for any indirect, incidental, special, consequential, or punitive damages, or for any loss arising from reliance on the assessment, a report, a score, or the educational content on this page. Nothing in this section limits liability that cannot be limited by law, including for death or personal injury caused by negligence, or for fraud. [Cap and carve-outs to be confirmed with counsel and aligned with the Terms of Service.]
Check our work
Verify any of this.
These statements have not been evaluated by the Food and Drug Administration. These products are not intended to diagnose, treat, cure, or prevent any disease. This assessment is educational and is not medical advice.
