// tool reference · Governance & Advisory
PII Classifier Free
Find and classify personal data in a text blob.
What it does
Detects and classifies personal data in a text sample — useful before you ship a dataset, a log export, or a support attachment.
When to use it
- Data-protection review; checking whether a log export is safe to share.
Inputs
Field names are the actual form parameters, verified against source.
| Field | Type | Required | Notes |
|---|---|---|---|
| text | textarea | required | The text to classify. |
What you get back
Detected PII by category with location and sensitivity.
Worked example
Input
Order #4471 -- Tanaka Ichiro, tanaka@example.com, 090-1234-5678 Card 4111-1111-1111-1111 exp 12/28, My Number 1234-5678-9012
Output (abridged)
PII DETECTED -- 5 items, 3 categories
CRITICAL My Number (kojin bango) 1234-5678-9012
Japan's APPI treats My Number as tokutei kojin joho, with
handling rules stricter than ordinary PII. It should almost
never appear in an order record at all.
CRITICAL Payment card 4111-...-1111
PCI DSS: storing a PAN pulls this system into scope. If this
is a log, your logging pipeline is now in PCI scope too.
HIGH Name + email + phone (direct identifiers)
IF THIS IS A LOG EXPORT: do not share it. Redact My Number and the PAN
before it leaves this machine.
How it works
Deterministic pattern classification. Offline.
Limits
Read live from the platform configuration.
| Rate limit | 30 requests / 60s (platform default) |
Limitations — what it does not do
Pattern-based. It finds structured identifiers well and free-text PII poorly -- a name in prose may be missed. Absence of findings is not clearance to share.
Privacy
Analysed in memory. Not stored.
Standards
GDPRJapan APPIISO 27701 concepts