Load File Redaction with anonym.plus

Clear PII from a load file while the field structure stays intact.

In simple terms, PII redaction is the on-device process of finding and masking personally identifiable information in a document before it is shared.

Load-file redaction is the removal of personal data from the metadata and extracted-text fields of an EDRM exchange index. The file is delimiter-bound. ASCII 20 separates fields, ASCII 254 quotes them, and ASCII 174 splits multi-value entries. anonym.plus holds that layout.

When this applies

A DAT or OPT index keys every document into the review platform. Inject a stray delimiter into one field and the row shifts, so pages bind to the wrong record. Extracted text that keeps a name the image no longer shows is a leak.

How anonym.plus handles it

  1. Open the DAT or CSV in anonym.plus on your device.
  2. It maps the delimited fields, including extracted text.
  3. The tool flags names, contacts, and IDs per field.
  4. Confirm the flags and keep the delimiters intact.
  5. Replace or mask each confirmed value.
  6. Save the clean index on your device.

What you need to provide

PII entity types detected

Categoryanonym.plus entity typeExample
NamesPERSONCustodian field → [CUSTODIAN]
ContactEMAIL_ADDRESSFrom field → [EMAIL]
DatesDATE_TIMEDateSent → [DATE]
Free textPERSONExtractedText name → [NAME]
IdentifiersUS_SSNtext SSN → [SSN]
Document IDsPERSONBegBates → [DOC_ID]

Compliance achieved

Anonymize load files offline — see plans & start free →

Limitations & cautions

These files are delimiter-sensitive. A stray quote or comma in a swapped value can break a field. Test the output in your platform on a small batch first, and keep Bates fields untouched unless you mean to change them.

Frequently asked questions

What breaks when a cut hits a load file?

A swapped value holding a delimiter shifts every field after it on that row. Pages then bind to the wrong document, and the volume can fail to import outright. The tool escapes values instead of writing raw delimiters.

Why must extracted text match the produced image?

Reviewers search the extracted text, not the picture. If the page is blacked out but the text still holds the name, the first keyword search defeats the cut. Clear both.

Do BegBates and OPT page pointers survive?

Yes, unless you ask otherwise. Bates and image-path fields key the whole exchange, so they stay in place while personal data in other fields is swapped.