← Back to pl3nish.com
DATA METHODOLOGY

How Pl3nish Calculates Your Nutrition

Transparency is a core value. This page documents exactly where our nutrition data comes from, how we process it, and why we believe it sets a higher standard than any other consumer nutrition app.

USDA FoodData CentralNutritionixFooDBPhenol-Explorer

Why We Publish Our Methodology

Most nutrition apps treat their data pipeline as a black box. You get a calorie number and a macro split, but no way to know where those numbers came from, what database was queried, or whether the values were estimated, crowdsourced, or verified.

Pl3nish takes the opposite approach. Every nutrition value in the app traces back to a specific, citable source — and this page shows you exactly how that data moves from raw databases to your daily dashboard. If you’re a journalist, researcher, dietitian, or simply someone who reads labels carefully, this is for you.

The Data Pipeline: 6 Steps from Database to Dashboard

01

Recipe Ingestion & Normalization

Every recipe enters our pipeline with a structured ingredient list. Each ingredient string is parsed into quantity, unit, and food item — then normalized to a canonical form (e.g., '2 cups baby spinach, packed' becomes 'spinach, baby, raw' at 60g).

We maintain a curated mapping table of ~3,000 ingredient aliases to canonical USDA food names, covering brand variations, regional terms, and common misspellings.

02

USDA FoodData Central Matching

Each canonical ingredient is matched to its USDA FoodData Central (FDC) entry — primarily from the SR Legacy and Foundation Foods datasets, which contain laboratory-verified nutrient profiles.

Matching uses a dual-strategy approach: exact title matching on normalized names, plus token-based fuzzy matching for ingredients that don't have a direct canonical entry. Every match is validated against expected nutrient ranges.

Source: USDA FoodData Central
03

Per-Serving Nutrient Calculation

USDA provides nutrient values per 100g. We scale each nutrient by the actual gram weight of the ingredient in the recipe, sum across all ingredients, then divide by serving count — producing precise per-serving values for calories, protein, fat, carbohydrates, fiber, and 30+ micronutrients.

This is not estimation. It is arithmetic applied to laboratory data. A recipe with 12 ingredients produces 12 × 35+ individual nutrient calculations, all traceable back to specific FDC food IDs.

04

Polyphenol & Phytonutrient Enrichment

Matched ingredients are cross-referenced against FooDB v1.0 compound data and Phenol-Explorer, the two most comprehensive food phytochemical databases in existence.

FooDB (University of Alberta) covers 70,000+ chemical compounds across 1,000+ foods. Phenol-Explorer (INRA, France) provides quantitative polyphenol composition data from peer-reviewed chromatographic analyses. Together, they power the polyphenol tracking that no other consumer nutrition app offers.

Source: FooDB + Phenol-Explorer
05

Real-Time Food Lookups

For foods not in our recipe library — restaurant meals, branded products, barcoded items — the Nutritionix API provides instant, verified nutrition data from its database of 900+ restaurant chains and 1,000,000+ branded foods.

Nutritionix data is used for on-the-fly logging only and is clearly distinguished from our USDA-calculated recipe nutrition. Both data paths feed into the same daily tracking dashboard.

Source: Nutritionix
06

Quality Assurance & Auditing

Every enriched recipe passes through automated validation: calorie ranges, macronutrient ratios, and diet-tag consistency checks flag anomalies before data reaches users.

We run periodic full-library audits — checking for vegetarian recipes with meat ingredients, vegan recipes with dairy, and nutrition values that fall outside expected ranges. Flagged recipes are reviewed and corrected manually.

Data Sources at a Glance

USDA FoodData Central

Visit →
ROLE
Primary nutrition source for all recipe data
COVERAGE
600,000+ foods with laboratory-verified nutrient profiles (calories, macros, 30+ micronutrients)
DATASETS USED
SR Legacy, Foundation Foods
LICENSE
Public domain (U.S. Government work)

Nutritionix

Visit →
ROLE
Real-time food lookups for restaurants, branded items, and barcode scanning
COVERAGE
900+ restaurant chains, 1,000,000+ branded products, natural language food parsing
DATASETS USED
Nutritionix Verified Database
LICENSE
Commercial API (licensed)

FooDB

Visit →
ROLE
Food compound data for polyphenol and phytonutrient tracking
COVERAGE
70,000+ chemical compounds across 1,000+ food items, including flavonoids, anthocyanins, isoflavones, and lignans
DATASETS USED
FooDB v1.0 (2020-04-07)
LICENSE
Academic use — Wishart Research Group, University of Alberta

Phenol-Explorer

Visit →
ROLE
Quantitative polyphenol composition data with compound-level breakdowns
COVERAGE
500+ polyphenols across 400+ foods, sourced from peer-reviewed chromatographic analyses
DATASETS USED
Phenol-Explorer 3.6
LICENSE
Academic — INRA (French National Institute for Agricultural Research)

What We Don’t Do

No crowdsourced data
We don't accept user-submitted nutrition values. Every number traces to a verified source.
No AI-estimated nutrition
Calories and nutrients are calculated from measured database values, never generated or guessed by a language model.
No hidden rounding
Per-serving values are calculated to full precision. What you see is what the math produced.
No unattributed data
Every data source is named and linked. We show our work because we believe you deserve to verify it.

For Journalists, Researchers & Dietitians

If you’re covering nutrition technology, evaluating apps for clinical recommendations, or writing about food data quality, we welcome scrutiny. Our methodology is public because we believe transparency builds trust — and because the nutrition app industry needs more of it.

For press inquiries, data partnership requests, or technical questions about our pipeline, contact:

Casey Cahoon — Founder & CEO
casey@pl3nish.com
WCC Systems, LLC · Denver, Colorado

Data Disclaimer: Nutrition values in Pl3nish are calculated from third-party databases and are provided for general informational purposes only. Actual nutrient content may vary based on ingredient brand, preparation method, and natural food variation. Pl3nish is not a medical tool and is not intended to diagnose, treat, or manage any health condition. Always consult a qualified healthcare professional for medical nutrition advice.

Attribution: Recipe nutrition data sourced from USDA FoodData Central (public domain). Polyphenol data: FooDB v1.0 (2020-04-07) — foodb.ca. Polyphenol compound breakdowns: Phenol-Explorer 3.6 — phenol-explorer.eu. Real-time food data powered by Nutritionix by Syndigo.

← Back to Pl3nish·Our Data Sources →