Sanitext

Italian PII Detection and Redaction

Sanitext detects and redacts PII in Italian text, including first and last names, emails, phone numbers, addresses, dates, URLs, account numbers, and secrets. Unlike English-only tools such as AWS Comprehend's default model, it reads Italian natively across 30+ languages from one API endpoint.

Most PII tools are tuned for English. Italian names, address formats like "Via Roma 12", and phone numbers with the +39 prefix slip through or get mislabeled. Sanitext was trained on 30+ languages, so it reads Italian text the way it reads English. You send Italian text to one endpoint and get back labeled spans or redacted output, with no separate language model to configure.

Italian PII types Sanitext detects

LabelItalian exampleNotes
FIRSTNAMELuca, GiuliaItalian given names
LASTNAMERossi, EspositoItalian family names
PREFIXDott., Avv., Ing.Titles before a name
EMAILluca@esempio.itAny domain, including .it
PHONE+39 320 123 4567With or without +39
ADDRESSVia Roma 12, 20100 MilanoStreet, CAP, city
DATE15/03/1990, 3 marzo 2024Numeric and written months
ACCOUNTIT60X0542811101000000123456IBAN-style and card numbers
SECRETpf_live_3f8...API keys and tokens

bash

curl -X POST https://api.sanitext.app/v1/redact \
  -H "Authorization: Bearer $SANITEXT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Salve, sono Luca Rossi, scrivetemi a luca@esempio.it o al +39 320 123 4567."
  }'

# Response
# {
#   "text": "Salve, sono [FIRSTNAME] [LASTNAME], scrivetemi a [EMAIL] o al [PHONE]."
# }

FAQ

Does Sanitext detect PII in Italian without extra configuration?+

Yes. You send Italian text to the same /v1/detect or /v1/redact endpoint you use for English. The model reads 30+ languages, so there is no language flag to set and no separate Italian model to deploy.

Can Sanitext detect Italian phone numbers with the +39 prefix?+

Yes. Sanitext detects Italian mobile and landline numbers with or without the +39 country code, such as +39 320 123 4567. It labels them PHONE and returns the byte offsets so you can mask or extract them.

Does Sanitext detect codice fiscale or IBAN in Italian text?+

Sanitext detects structured account and card numbers, including IBAN-style strings, under the ACCOUNT label. Codice fiscale is best caught when it appears in a clear pattern. Always review output, since Sanitext is a data-minimization aid, not a guarantee.

Is my Italian text sent to a third-party LLM?+

No. Sanitext runs an open-weights model on our own Cloudflare infrastructure. Your text never reaches an external LLM, and we store no raw text. We log only character counts and timings, which helps with EU and GDPR work.

Redact Italian PII in 60 seconds

Sign up free and get 300K characters with no card. Grab an API key and run your first Italian redaction with a single curl command. See the docs or test it live in the playground.

Get your free API key