Handwritten Receipt Detection

When enabled, Handwritten Receipt Detection classifies whether a receipt appears to contain handwritten content and returns isHandwritten plus handwrittenScore. Route the result according to your workflow rather than inferring which internal feature triggered it.

Availability

📘

Availability

Handwritten Receipt Detection is part of the fraud detection suite, included on the Advanced plan and above. Enable Handwriting Detection in your account Feature Settings when it is available on your plan. Once enabled, the check is attempted automatically on eligible verbose requests; a populated result is not guaranteed. See Plans & Feature Access.

📘

Extraction endpoints only

Handwritten Receipt Detection is currently available on Data Extraction endpoints, not Receipt Validation endpoints. If you need it in a validation workflow, contact [email protected] to discuss your use case.

How It Works

The check returns a boolean recommendation and a numeric score. Internal techniques may change and are not part of the public API contract; build integrations on the documented response fields.

Response

entities.handwritingDetection, its data object, and the flag and score properties are optional and may be omitted. Feature enablement does not guarantee a populated result for every document. Check for a usable boolean or score before applying a rule.

Example input

A receipt containing clearly handwritten transaction details is a useful illustration of the input type. Handwriting is not proof of fraud; use the returned result as a workflow signal.

Illustrative receipt containing handwritten content; handwriting alone is not proof of fraud. The response below is a separate schema example and is not attributed to this image.

Example response

The response below illustrates the documented verbose-extraction shape. It is a schema example, not a promised score for every document containing handwriting:

{
  "entities": {
    "handwritingDetection": {
      "data": {
        "isHandwritten": true,
        "handwrittenScore": 0.99
      }
    }
  }
}

Best Practices

• Build rules on isHandwritten — it's Taggun's tuned recommendation. Use handwrittenScore when you need to tighten or loosen routing for your use case (e.g. review a wider band before rejecting).
• Combine with the other fraud detection checks for layered protection.
• Give submitters guidance on clear, legible handwriting — see End-User Capture Guidance.

Excluding Handwritten Digits from Extraction

To exclude detected handwritten digits from extracted values, enable Remove Handwritten Digits in your account Feature Settings when the setting is available on your plan. This setting changes extraction behaviour rather than adding a dedicated response field. See Plans & Feature Access.

FAQ

Will this detect all handwriting on a receipt?
Detection depends on the submitted document and is not a guarantee that every handwritten element will be identified. Use the flag as a workflow signal and validate any score threshold on your own labelled data.

Need something not covered here? Contact [email protected] — see Contacting Support.