PARSER GUIDES

Disable AI Extraction

Fields & TablesField Setup

Turn off AI extraction for specific fields

By default, PARSER uses AI to extract each field from your documents. You can turn off extraction for specific fields. When a field has AI extraction disabled, it won't be extracted from the document and must be filled another way (e.g. manually or via Post Processing).

Why disable AI extraction?

• You fill it yourself: You always type or paste the value (e.g. a client code you assign).
• The uploader fills it: On Standalone Upload Pages, you can show selected fields before upload (e.g. student ID or customer email).
• Post Processing fills it: You use lookups, custom code, or other Post Processing to set the value based on other fields.
• The data isn't in the document: The field exists for organization or export, but the source is elsewhere (e.g. from another system).
• Extraction is unreliable: The AI often gets it wrong for this field, and you prefer to enter or correct it yourself.

How to disable AI extraction

1. Go to Settings -> Fields & Tables (or Fields & Tables in the parser sidebar).

2. Click the field you want to change to expand it.

3. Find Enable AI extraction.

4. Turn the switch off to disable extraction for that field.

5. Click Save Fields.

When Enable AI extraction is off:

• The field is not extracted from the document.
• The field can still appear in the data table and in exports; you (or Post Processing) must provide the value.
• You can edit the value manually in the document view or after export.
• If the field is visible, top-level, and not a table field, you can add it to a Standalone Upload Page so uploaders fill it before processing.

Enabling AI extraction again

To start extracting the field again:

1. Open Fields & Tables and expand the field.

2. Turn Enable AI extraction on.

3. Click Save Fields.

New documents will then have the field extracted again. For existing documents, use Re-process document if you want them to match.

Using with Post Processing

When a field has AI extraction disabled, you can use Post Processing to fill it automatically, for example:

• Lookups: Match another field (e.g. Vendor Name) to a CSV file or parser-data source and set this field (e.g. Vendor ID).
• Custom code: Use JavaScript to compute the value from other fields or rules.

See Post Processing for an overview.

Using with Standalone Upload Pages

For fields that should be provided by the person uploading the document, add the field to your Standalone Upload Page:

1. Go to Integrations -> Upload Page.

2. Open Show Customization Options.

3. Enable Fields shown before upload and select the field.

The uploader fills the value below the upload area before processing starts. PARSER stores that value with the document and applies it to extracted data before workflows, post-processing, exports, and the optional review step. If AI extraction is still enabled for that field, the upload-page value overrides the extracted value for that upload.

See Standalone Upload Pages for setup details.

Tips

• Use disabling when you do not want AI extraction for that field. Use Fields shown before upload when upload-page visitors should provide a value while other upload methods still use extraction.
• If you disable extraction, add a short AI Description when you turn it back on--it can help the AI extract more accurately.

Next steps

• Add/Edit Fields - Add fields and configure extraction.
• Post Processing -> Lookups - Fill fields from CSV files or parser data.
• Post Processing -> Custom Code - Use custom logic to set field values.