Products
Technology
Solutions
Resources
Company
Open AppTalk to Sales
Language
Home/How it works
How it works

From raw food data to real intelligence.

Barliva is a five-stage pipeline. Each stage adds structure and meaning, turning scattered public records and manufacturer feeds into a canonical, multilingual food knowledge graph with explainable health intelligence.

01
Ingest
Many sources
02
AI normalize
Clean & classify
03
Knowledge graph
Canonical entities
04
Health intelligence
Scores & flags
05
Consumer insight
Plain language
Stage 01

Ingest from many sources

We pull from OpenFoodFacts, manufacturer data, product metadata and other structured feeds. No single source is treated as ground truth — each is evidence to be reconciled. See data sources →

Stage 02

AI normalization

Free-text fields are cleaned, parsed and translated. Ingredients are classified, additives mapped to E-numbers, allergens detected, and conflicting values reconciled with confidence scoring and audit trails.

Stage 03

Canonical knowledge graph

Records resolve into canonical entities — one product, one ingredient, one allergen — connected by relations and expressed across languages. Inside the engine →

Stage 04

Health intelligence

A category-aware baseline plus transparent penalties and capped bonuses produces an explainable 0–100 health score, with additive and allergen context. How scoring works →

Stage 05

Consumer insight

Finally, Barliva generates localized, plain-language explanations so anyone can understand what a product contains and why it scored the way it did.

See it live

Watch the pipeline in action.

Scan a product in the app and see every stage's output.