Local Files (loaded into Postgres at seed time)
| File | Records | Notes |
| Data/EAFUS - FoodSubstances.csv |
3,984 |
FDA substances added to food (3,985 lines − 1 header) |
| Data/foodb_2020_04_07_json/Compound.json |
70,477 |
Core FooDB compound records |
| Data/foodb_2020_04_07_json/CompoundSynonym.json |
171,240 |
Synonyms per compound |
| Data/foodb_2020_04_07_json/CompoundsHealthEffect.json |
11,062 |
Compound × health effect links |
| Data/foodb_2020_04_07_json/HealthEffect.json |
1,435 |
Unique health effect descriptors |
| Data/foodb_2020_04_07_json/Food.json |
992 |
Food sources (e.g. apple, beef, green tea) |
| Data/foodb_2020_04_07_json/CompoundsFlavor.json |
11,775 |
Compound × flavour descriptor links |
| Data/foodb_2020_04_07_json/Flavor.json |
883 |
Unique flavour descriptors |
| Data/foodb_2020_04_07_json/Content.json |
~5,600,000 |
Compound concentrations in foods — skipped at seed (too large) |
| Local total (excl. Content.json) |
~269,000 |
|
Live APIs (fetched at seed / monthly refresh)
| Source | Records | Notes |
| EU Food Additives |
13,849 |
EC Datalake API v2.0 — one row per E-number × food category |
| EU Flavourings |
3,579 |
EC Datalake API v2.0 |
| EU Health Claims |
3,342 |
EC Datalake API v2.0 — 1,022 authorised + 2,320 not authorised |
| EU Novel Foods |
955 |
EC Datalake API v2.0 — aggregated from 6,516 raw rows |
| UK FSA Food Additives |
999 |
data.food.gov.uk — 333 unique E-numbers × 3 jurisdictions |
| UK FSA Smoke Flavourings |
~50–200 |
data.food.gov.uk |
| API total |
~22,700 |
|
Hardcoded reference data
| Source | Records | Notes |
| Allergens |
16 |
EU Regulation 1169/2011 Annex II (14) + 2 US-only (FASTER Act 2021) |
Also queried live during each session
| Source | Access |
| Open Food Facts | world.openfoodfacts.org API v2 — product ingredient lists & nutritional data |
| FlavorDB2 | cosylab.iiitd.edu.in/flavordb2 — molecular flavour pairing data |
| PubMed / NCBI Entrez | eutils.ncbi.nlm.nih.gov — scientific literature search |
| USDA FoodData Central | api.nal.usda.gov/fdc — nutritional profiles |
| Tavily Web Search | Real-time web search for market intelligence, trends, pricing |
Grand total loaded into DB: ~292,000 records
+ ~250,000–300,000 generated ingredient aliases