Skip to content

Choosing your edition

Both editions are the same product (interface, accounts, providers, Docker). They differ on one thing only: the detection engine.

Regex-based engine, no heavy dependency. It covers deterministic data: IPs, emails, API keys and secrets, IBANs, payment cards, French NIR, phone numbers, plus names preceded by a civility title (“Mr. Dupont”). Free, public code, installs in seconds.

What it does not do: detect people, client, or organization names in free text, without a title. That is the structural limit of a regex approach.

Adds machine-learning detection: Microsoft Presidio and spaCy NER (French and English), CamelCase, User-Agent and query-param heuristics, key-driven detection on structured data (JSON, key-value). It is the edition measured at 0% leakage across 3,340+ values, prose names included.

It also adds: document anonymization (Word, Excel, PDF, OCR) with redacted copies, LDAP / Active Directory, multi-tenant seat management, a commercial license and support.

Evaluate with Community: it is enough to audit the code and validate the workflow. Move to Enterprise when the data you protect includes names in free text (a client name buried in a stack trace, an organization in a ticket), which is the case in virtually every real log.

Contact: contact@tokenveil.eu or request a demo.