User information review
Users directly enter information that may affect food choices, such as allergies, health concerns, current medications, supplements, and eating habits.
Information
AI explanation structure based on health and current medication information
Can I Eat? is not limited to barcode scan results. The app compares user-provided health information and current medication information with food data to organize caution-focused reference information.
AI engine flow
Inside the app, explanations are created in this order.
Product data organization
Barcode, product name, ingredients, nutrition facts, allergy labels, and country data pack information are normalized into the same review structure.
Reference source matching
Public drug labels, food recall data, supplement ingredient references, and country-specific food safety sources are checked for relevant items.
AI explanation generation
The on-device AI analyzes the user's current status, current medications, and the food found through barcode search, then organizes the result into language that is easy to understand. Instead of making a definitive diagnosis, it separates caution reasons, ingredients to check, and situations where professional consultation is needed.
Detailed structure
Explanations pass through several layers.
Input layer
User health information, current medication information, product data, and country data packs are stored separately and combined only when needed for analysis.
Normalization layer
Product names, ingredient names, medication names, and nutrition units are organized into comparable formats.
Evidence search layer
Public sources such as FDA, NLM, NIH, MFDS, Food Safety Korea, and USDA FSIS are used as reference points.
Caution analysis layer
Ingredients, allergies, medication information, recall history, potential overconsumption, and user-provided information are compared together.
Explanation layer
The result is written as reference information such as caution needed, additional review, or professional consultation recommended.
Safety notice layer
The app clearly states that AI can make mistakes and that medication or disease-related decisions require professional consultation.
AI can make mistakes. These explanations are reference information for food choices, not diagnosis, prescription, or treatment. If a decision involves medication or a health condition, consult a licensed medical professional or pharmacist.
Sources
We disclose the reference sources we use.
The sources below are public references used now or planned for future data pack construction. Open each row to see how it is used.
FDAopenFDA API
A starting point for public FDA data APIs covering drugs, foods, and medical devices.
openFDA API
A starting point for public FDA data APIs covering drugs, foods, and medical devices.
Used as a reference source for publicly available drug labels, adverse event data, and food enforcement information.
Open source pageFDAopenFDA Drug Labeling
Public API for prescription and over-the-counter drug label information.
openFDA Drug Labeling
Public API for prescription and over-the-counter drug label information.
Helps review warnings, adverse reactions, ingredients, and usage information when medication-related explanations are needed.
Open source pageFDAopenFDA Drug Adverse Event
Public API based on FDA adverse event reporting data.
openFDA Drug Adverse Event
Public API based on FDA adverse event reporting data.
Used carefully as report data, not proof of causation, when checking whether reports exist for a drug or ingredient.
Open source pageFDAopenFDA Food Enforcement Reports
Public API for FDA food enforcement and recall data.
openFDA Food Enforcement Reports
Public API for FDA food enforcement and recall data.
Used to review recall, safety alert, and enforcement information related to products or ingredients.
Open source pageNLMDailyMed
A drug label database provided by the US National Library of Medicine.
DailyMed
A drug label database provided by the US National Library of Medicine.
Important for reviewing current medication information alongside food ingredients.
Open source pageNIHNIH Office of Dietary Supplements
Scientific fact sheets for vitamins, minerals, herbs, botanicals, and supplements.
NIH Office of Dietary Supplements
Scientific fact sheets for vitamins, minerals, herbs, botanicals, and supplements.
Used to reduce exaggerated supplement language and distinguish consumer-facing and professional reference information.
Open source pageUSDAUSDA FSIS Recall API
Recall and public health alert data for meat, poultry, and egg products under USDA FSIS.
USDA FSIS Recall API
Recall and public health alert data for meat, poultry, and egg products under USDA FSIS.
Complements FDA sources for US food data packs when food safety alerts fall under FSIS jurisdiction.
Open source pageMFDSFood Safety Korea Public Data
Public data service from Food Safety Korea, operated by Korea's Ministry of Food and Drug Safety.
Food Safety Korea Public Data
Public data service from Food Safety Korea, operated by Korea's Ministry of Food and Drug Safety.
Used for Korean data packs covering food safety, product, recall, and labeling information.
Open source pageMFDSKorea Food and Drug Data Portal
MFDS public data guide for food, drug, and medical device datasets.
Korea Food and Drug Data Portal
MFDS public data guide for food, drug, and medical device datasets.
Used to review Korean public data structures and API availability.
Open source pageMFDSKorea Drug Information Search
Korean drug product information search provided through MFDS systems.
Korea Drug Information Search
Korean drug product information search provided through MFDS systems.
Used to compare user-entered medication names with approved Korean drug information.
Open source page