Food label scanner

A food label scanner that keeps the uncertainty.

Aurascan organizes what a packaged-food label says, adds source-backed context, and keeps facts, interpretation, and unknowns visibly separate. It does not turn a barcode into a diagnosis or a magic safety score.

See the sample result Read the methodology
  • Sources stay visible
  • Barcode, not camera footage

Crawlable sample result

A sample that says what it is

This is an illustrative interface example built from a constructed four-ingredient label. It is not a live product analysis, health verdict, endorsement, or substitute for the sources shown in a real result.

The point is the contract: readers can distinguish extracted label facts from context and from what cannot be inferred.

Illustrative sample · not a live scan

Example pantry label

4 ingredients

Rolled oats

Read directly from the constructed example label.

Label fact

Almonds

Read directly and carried into the example allergen notice.

Label fact

Cane sugar

The name is visible; the ingredient list alone does not provide a serving amount.

Needs context

Sea salt

Ingredient order does not establish dose or an individual health effect.

Limit

Label notice: contains almonds.

Unresolved: serving amount is not provided.

Method

From barcode to an inspectable result

  1. STEP 1

    Decode the barcode

    The camera reads the barcode on the device. The decoded number starts the product lookup.

  2. STEP 2

    Retrieve the label

    Aurascan combines available product records and keeps provenance for the fields that contributed.

  3. STEP 3

    Separate fact from interpretation

    Label text stays distinct from regulatory context, research context, and unresolved questions.

  4. STEP 4

    Show sources and limits

    The result surfaces citations, evidence strength, uncertainty, and the methodology version behind it.

How evidence is labelled

Label fact
Printed ingredients, allergens, nutrition fields, and product identity.
Regulatory context
Rules or assessments attributed to a named regulator or public authority.
Research context
Human, observational, or mechanistic evidence with its limits kept visible.
Unresolved
A gap stays a gap when the available label or evidence cannot support a conclusion.

Limits stay beside the result

  • A label can be incomplete, outdated, or different across markets and package sizes.
  • Ingredient order does not reveal exact dose, exposure, or an individual response.
  • Research context can describe evidence; it cannot diagnose, treat, or replace clinical advice.
  • AI-assisted research can be wrong. Aurascan exposes sources and unresolved states so claims can be checked.
Inspect the full methodology

Scan privacy

The camera reads locally

When barcode scanning is enabled, the camera decodes the number on the device. Only the decoded barcode is sent for product lookup; camera footage is not uploaded as part of that barcode scan.

Account and scan-history handling is described in the privacy policy, including choices for deletion and data requests.

Read the privacy policy

Private beta

Bring your own label

Join the beta waitlist to scan packaged-food labels and inspect the sources, evidence states, and limitations behind each result.

Join the beta waitlist

Educational context only. Aurascan does not provide medical advice.