NOVYRA LLC

Strata scoring methodology and independence

Product: Strata — Supplement Scanner

Company: NOVYRA LLC

Methodology version: 2.2.0

Last updated: September 21, 2026

This page explains how Strata scores products, what those scores mean, and how NOVYRA LLC keeps scoring independent of brands.


Our core commitment: independence

Strata exists to help people make clearer choices about supplements and selected foods. That only works if the score cannot be bought.

Revenue is intended to come from people who choose Strata Pro — not from brands purchasing influence.


What a Strata score is

A Strata score is an educational opinion on a 0–100 scale, produced by applying published rubric rules to available label data, public databases, and Strata’s reviewed evidence registry.

It is not:

Always talk with a qualified healthcare professional before starting, stopping, or changing supplements or medications — especially if you are pregnant, nursing, taking medication, or managing a medical condition.


How scoring works (version 2.2.0)

The total is a transparent Strata rubric. Numerical weights are product-design choices, not research effect sizes.

Starting point

Every assessment starts at 55 — a rubric anchor, not evidence of quality.

Insufficient label data (v2.2)

When a supplement label lists ingredients but provides no usable doses, or coverage is too thin to support a confident grade, Strata marks the result as Insufficient label data.

This stops thinly labeled products from looking like a quiet B just because little negative data was available.

What can move the score

Strata combines:

  1. Nutrition rubric (foods) — per-serving values only (never silently swapping in per-100 g). Examples of published bands include sugar, protein, fiber, sodium, and NOVA group adjustments for food only. Missing nutrition values receive no adjustment.
  2. Reviewed ingredient evidence — only citation-backed claims can add or subtract research-related points. Legacy uncited rules do not masquerade as reviewed research.
  3. Dose context — an ingredient cannot earn efficacy-style points when its amount is unavailable. Disclosed amounts are compared with reviewed reference ranges when appropriate. Mass units are normalized; empty, negative, non-finite, or unsupported units never earn dose-disclosure points.
  4. Transparency — points for usable disclosed amounts; proprietary-blend / artificial-flavor style penalties are labelled as transparency rules, not clinical evidence. Organic / non-GMO / palm-oil-free style marketing labels do not earn health bonuses by themselves. Ingredient or additive counts alone do not create penalties.
  5. Referenced-limit flags — exceeding a reviewed intake limit can cap the educational score and prompt professional review. That cap is a conservative educational flag, not an individual safety diagnosis.

Ingredient adjustments are bounded (supplements roughly −30 to +24; foods roughly −26 to +18). The total is clamped to 0–100. Component values are designed to sum to the displayed score, including bounds and caps.

Evidence Passports

For reviewed ingredients, Strata can show an Evidence Passport: claim, evidence grade, population, dose context, limitations, citations, and review date. Scientific evidence strength and label-source confidence are reported separately on purpose.

Search vs detail

A preliminary search estimate is labelled as preliminary until the shared product assessment resolves. Search and detail are meant to converge on the same versioned report for the same canonical product identity.

Audits and disputes

Scores can carry a derivation / audit trail (inputs, rule IDs, point deltas, sources). If you believe the underlying data is wrong, use the in-app correction / dispute option so it can be reviewed.


Data sources we may use

Depending on the product, Strata may draw on:

Availability varies by barcode, catalog coverage, and source quality. A missing or incomplete label limits what the rubric can responsibly conclude.


AI in Strata

Some explanations and chat features use AI to put results in plain language. AI commentary is labelled as analysis / opinion support. It does not silently override the versioned numeric rubric. When AI is involved as a source in an audit trail, it is marked distinctly from label data and reviewed research.


Versioning

FieldValue
Methodology version2.2.0
In-app constantStrataLegal.scoringVersion

When the rubric changes in a way that would alter published scores, we bump the methodology version so older audits remain interpretable.


Company

Strata is published by NOVYRA LLC (Florida).

Web: https://novyra.cc

For product support and corrections, use the in-app flows. For company contact, see novyra.cc.


Short version

Independent educational scores. Same published rules for every product in a category. Incomplete labels are marked provisional. No paying for a better grade. Not medical advice.

Source of truth in the app: methodology version lives in StrataLegal.scoringVersion (currently 2.2.0).

← Back to NOVYRA