How Feny Verify Works

Feny Verify is a rule-based classifier, not an AI. Every verdict traces back to a named, auditable source. Here is exactly what we check against.

Is Feny Verify an AI?

No. Feny Verify is a rule-based classifier. It matches product ingredients against curated dictionaries built from public regulatory and advocacy references. It does not generate, infer, or guess verdicts. The same product scanned twice returns the same result.

We look it up. We don't guess.

Feny Verify is a rule-based classifier, not an AI. Every verdict traces back to a named, auditable source. Every verdict is traceable. We do not infer or generate answers. When you scan a product, we match its ingredients against curated dictionaries built from public regulatory and advocacy references, and the result is reproducible, every time.

What we check against

| Dictionary | Entries | Languages | | --- | --- | --- | | Animal-derived terms | 330 | EN, FR, DE, ES, IT, NL, RO | | Ambiguous terms | 81 | EN, FR, DE, ES, IT, NL, RO | | E-numbers | 18 | Code-based | | Allergens (UK/EU/US) | 38 | Multilingual |

Our sources

  • Vegan Society guidance
  • EU Regulation 1169/2011 (FIC)
  • US FDA Big-9 allergen list
  • EU food additive register (EFSA)
  • Open Food Facts

Product data

All product lookups use Open Food Facts, an open and auditable database of 3 million+ products. Every result links back to the original Open Food Facts entry so users can check the source themselves.

This is why Feny Verify says "Not Vegan" (not "probably fine")

The classifier is not guessing. It is reporting a match between an ingredient on the label and a known animal-derived term in the dictionary. If the dictionary says the ingredient is animal-derived, the verdict is "Not Vegan". There is no confidence interval, no probability, and no override. The label says what it says. The dictionary says what it says. The verdict follows.

---

Dictionary version 1.1 · Last updated May 2026 · Updated quarterly.