Kanzura Methodology
Kanzura combines structured ingredient data, safety signals, and AI interpretation. Results are guidance-oriented and should be reviewed with professional judgment for high-stakes regulatory or medical decisions.
How scores are generated
- Ingredient identity is normalized using INCI/slug matching.
- Safety score is mapped from available source fields and capped to display range.
- Function/category labels are inferred from source taxonomy.
- When live data is unavailable, the system may fall back to curated catalog records.
Source and freshness
- Each search response includes a source tag: Supabase live or Catalog fallback.
- Each response includes a refreshed timestamp shown in the UI.
- Evidence and source coverage will expand progressively across product modules.
Need deeper validation or enterprise-grade documentation? Contact us for a source-level evidence package and compliance-ready reporting.