Untainted
Interactive Demo

Try the Intelligence

See how Untainted analyzes products in real-time against complex, natural-language profiles.

Intelligent Profile Builder

Describe your dietary needs in plain English, and we'll analyze products against your preferences.

Diet
No diet specified
Health
No conditions
Allergies
No allergies
Avoid
No avoidances

Type your preferences above and click Build Profile

Select a Product

Click on any product above to analyze it against your profile instantly.

The Food Intelligence Engine

From Labelling to Logic.

Untainted transforms unstructured packaging data into a structured food ontology. We combine computer vision, NLP, and regulatory rules to deliver a definitive safety verdict.

1. Data Ingestion

Our engine is agnostic to the input source. We ingest product data through three primary automated channels.

Barcode Lookup

Instant retrieval from our indexed database of 200,000+ Indian products (GS1/EAN standards).

OCR Image Scan

Vision models trained on 50k+ curved/crinkled packaging labels extracts Ingredient lists with 99.8% accuracy.

Direct Text

Raw text processing via API for R&D teams, recipe analysis, or supply chain auditing.

2. The Analysis Pipeline

Once data is ingested, it passes through our core processing stages: Normalization, Classification, and Evaluation.

Stage 1: Normalization

We convert messy, non-standard text into canonical IDs.

  • Synonym Mapping

    "Vit B1", "Thiamin", and "Thiamine Mononitrate" → en:thiamine

  • Code Decoding

    "INS 211", "E211", "Preservative (211)" → en:sodium-benzoate

Input Raw Text
"Ingredients: Refined Wheat Flour, Edible Veg Oil (Palm), E322, Sugar"
Output Normalized IDs
[
  "en:wheat-flour",
  "en:palm-oil",
  "en:lecithin",
  "en:sugar"
]
en:plant-based-food
en:cereal
en:wheat-flour
└─ contains: gluten
└─ type: grain
en:additive
en:sodium-benzoate
└─ risk: high
└─ origin: synthetic

Stage 2: Classification

We map normalized IDs to a hierarchical food ontology to understand relationships and properties.

Our database understands that wheat-flour is a child of cereal, which contains gluten.

This inheritance model allows us to flag "Hidden Allergens" even if the specific allergen word isn't explicitly on the label.

Stage 3: Evaluation

The final verdict. We cross-reference product data against the user's specific health profile and regulatory standards.

  • Rule Engine

    Checks for conflicts (e.g. Diabetics → Sugar > 5g/100g).

  • FSSAI/FDA Compliance

    Validates label claims like "No Added Sugar" against the actual ingredient list.

USER: JAIN + DIABETICNOT SAFE

Egg Content

Product contains Egg Powder. Conflicts with Jain (Vegetarian).

High Sugar

12g Sugar/100g exceeds your threshold of 5g/100g.

4. The Output

The result of this pipeline is a rich JSON payload delivered in under 50ms, ready for your application.

API Response Example
{
  "product": "Multigrain Cookies",
  "verdict": "NOT_SAFE",
  "confidence": 0.98,
  "flagged_ingredients": [
    "Refined Wheat Flour (Maida)",
    "Butter"
  ],
  "reasons": [
    {
      "ingredient": "Refined Wheat Flour (Maida)",
      "category": "refined_grain",
      "conflicts_with": ["no_maida", "gluten_free"]
    },
    {
      "ingredient": "Butter",
      "category": "dairy",
      "conflicts_with": ["dairy_free", "vegan"]
    }
  ],
  "safe_alternatives": ["Ragi Cookies", "Oats Digestive"],
  "processing_time_ms": 42
}

Powered by Massive Intelligence

Our rules engine is built on one of the world's most comprehensive ingredient databases.

93,000+
Unique Ingredients

Indexed and classified from raw manufacturing data.

26,000+
Additives Analyzed

Including full INS/E-number decoding and risk profiles.

50ms
Average Latency

Real-time analysis optimized for retail checkout speeds.

Experience the Intelligence

Whether you're building a platform or shopping for your family, Untainted provides the clarity you need.

For BusinessFor You