Methodology

How Lychee turns ingredient labels into useful context.

Lychee combines product lookup, label reading, ingredient normalization, public-source checks, and plain-language explanations. The goal is clarity, not certainty where the evidence does not support certainty.

Review sources Read FAQ
Core Workflow

The scan is built in layers.

1. Product match

Lychee first tries to match the barcode or product details to available product data. Coverage varies by region, product category, and how recently a formula was updated.

2. Ingredient parsing

Ingredient text is parsed into recognizable names, synonyms, additive codes, and cosmetic ingredient names where possible. Unclear or missing text should be checked against the package.

3. Source mapping

Ingredients are compared with public references such as food additive resources, nutrition data, cosmetic ingredient databases, and regulator pages.

Food Scores

Food analysis emphasizes ingredients, additives, processing cues, and label context.

For food, Lychee highlights ingredient-list signals such as additive categories, sweeteners, colorants, preservatives, common allergens, and processing indicators. Where a food score appears, it should be read as a summary of label-based signals rather than a full nutrition diagnosis.

A food can be acceptable for one person and inappropriate for another. Allergies, medical conditions, total diet, serving size, and age can change the practical meaning of a scan.

Cosmetic Scans

Cosmetic analysis separates ingredient concern from personal risk.

For skincare, makeup, haircare, deodorant, fragrance, and other beauty products, Lychee can explain ingredient roles and flag common concern categories such as fragrance terms, preservatives, colorants, exfoliating acids, UV filters, and potential allergens or irritants.

Cosmetic risk depends heavily on concentration, formulation, exposure, frequency of use, whether the product is rinse-off or leave-on, and a person's skin history. Lychee does not diagnose allergies, acne, dermatitis, pregnancy safety, or medication interactions.

AI Assistance

AI can explain labels, but sources and labels stay primary.

Lychee may use AI-assisted interpretation to make ingredient lists easier to read. AI output is treated as an explanation layer, not as an authority. The product package, public source references, and professional medical advice remain more important than a generated summary.

Reviewed May 15, 2026

This methodology page describes Lychee's public information approach. Source pages and regulations can change, so Lychee pages link to primary references where practical.