Discover how we protect your documents

Features

Automatic anonymization of personal data with over 95% accuracy. See the result with your own eyes.

Automatic Detection

Data types

Personal identification

Spanish tax numbers for individuals and legal entities (NIF, CIF), NIE, passport and Social Security affiliation number, with check-digit validation: the system verifies the identifier is mathematically valid before replacing it.

Names and legal entities

Full names with their forms of address and abbreviations, plus company and organisation names. Coreference resolution groups every mention of the same person and applies a single decision to all of them.

Contact details

Home addresses, landline and mobile numbers, email addresses, vehicle registration plates in current and historical formats, and websites.

Financial data

IBAN validated against ISO 13616 and monetary amounts in any currency. Both can be switched off when the amount is not personal data.

Administrative codes

Código Seguro de Verificación with a pattern configurable per organisation, and references to laws, decrees and circulars, recognised in order to exclude them and not destroy the document's legal information.

QR codes and barcodes

A printed QR code or barcode can contain the case file number or a person's identifier. A tool that only reads text leaves them untouched; anonimiza.do locates and handles them.

Relationships and roles

«Sister of», «husband of», «mayor of», «teacher at». A document identifies a person without naming them, and those links are generalised, neutralised or removed according to your configuration.

Your own data types

Describe the data type you need in plain language —medical record number, policy number, case file code— and the system finds it with no retraining and no regular expressions.

Multi-format with OCR

Compatibility

Native PDFs

Processes PDF documents with selectable text, extracting and anonymizing content while preserving the original structure and format.

Word (DOCX)

Anonymizes Microsoft Word documents while maintaining formatting, styles, tables and the structure of the original file.

Images (JPG, PNG)

Built-in OCR extracts text from images and document photographs to anonymize it automatically.

Scanned PDFs

Converts scanned PDFs (images) to text via OCR before applying detection and anonymization.

Batch processing

Upload and process multiple documents simultaneously. Anonymize entire folders of files in minutes.

Download in original format

Receive the anonymized document in the same format you uploaded: PDF stays PDF, DOCX stays DOCX.

Spreadsheets

XLS and XLSX files processed cell by cell, sheet by sheet. The data type is inferred from the whole column, so outliers get the same treatment as the rest.

Archives

Upload a ZIP and the system unpacks it, handles each file according to its type and returns the whole set compressed, with no downloading document by document.

Security & Privacy

Protection

Encryption in transit and at rest

All file transfers use TLS 1.3, so data never travels as plain text, and stored documents are encrypted with AES-256, the standard used by governments and financial organizations worldwide.

Metadata removal

Every anonymized document is delivered free of metadata: we strip the file’s hidden information (author, software, dates or geolocation) so no identifiable trace remains beyond the content.

EU-based servers

All infrastructure is located in European data centers, guaranteeing your data never leaves EU territory.

Configurable auto-deletion

Define the retention period for your documents (24h, 48h, 72h, 7 days or custom). Irreversibly deleted when the period expires.

GDPR compliant

The tool facilitates GDPR compliance, allowing you to anonymize personal data systematically and auditably.

No access to your content

We never access the content of your documents. Processing is automated and files are deleted according to configured retention.

ENS Medium certification

CEDESA DIGITAL, S.L. holds certification of conformity with the Esquema Nacional de Seguridad (Spain's National Security Framework) in MEDIUM category, under Royal Decree 311/2022, on infrastructure certified in the HIGH category.

Full activity log

Access, processing, review, approval, download, pseudonym reversal and configuration changes are all recorded with user, date and time.

One configuration for each document type

Configuration profiles

Every document needs different rules. Create custom profiles and reuse them with one click.

Configuration
Full name
ID / NIE
Email
IBAN
Phone
Address
Date of birth
Social Security #
Preview
Full name [FULL_NAME]
ID / NIE [ID_NUMBER]
Email maria@company.com
IBAN [IBAN]
Phone [PHONE]
Address [ADDRESS]
Date of birth 15/03/1985
Social Security # 28 12345678 09

We identify every sensitive piece of data

Smart detection

Twenty predefined data types in seven families, plus the ones you define yourself. The engine chains recognisers with check-digit validation, Spanish entity models, coreference resolution and detection of family and role relationships.

lease_agreement.pdf

LEASE AGREEMENT

Madrid, March 15, 2026

Mr./Ms. María García López, of legal age, with ID number 12345678Z, residing at C/ Gran Vía 28, 3ºB, 28013 Madrid, phone +34 612 345 678, email maria@company.com, hereinafter referred to as "the tenant".

Bank details: ES91 2100 0418 4502 0005 1332

Born on 15/03/1985, Social Security number 28 12345678 09, vehicle license plate 1234 BCD.

Track the status of every document

Real-time tracking

From upload to download, every document has full traceability.

Uploaded
Queued
Processing
Completed
Downloaded
Document Date Status Action
📄 employment_contract.pdf Apr 8, 2026 Processing —
📄 payroll_march_2026.pdf Apr 7, 2026 Completed Download
📄 medical_report_anna.docx Apr 6, 2026 Downloaded View

Your documents are deleted automatically

Secure storage

Comply with the GDPR data minimisation principle effortlessly.

Document uploaded Deleted
0h 12h 24h 36h 48h
48h configured Irreversible

Configurable retention period

Choose 24h, 48h, 72h, 7 days or custom. You decide.

Irreversible deletion

Once the period expires, files are destroyed with no possibility of recovery.

Encrypted in transit and at rest

TLS 1.3 for transfers, AES-256 for storage.

Your documents are encrypted in transit and at rest, processed automatically with nobody reviewing their content, and deleted when you decide. CEDESA DIGITAL, S.L. acts as the processor and holds ENS Medium certification.

TLS 1.3 AES-256 ENS Medium Processor, Art. 28 GDPR GDPR

Choose what replaces each piece of data

Substitution modes

Detecting data is half the work. The other half is deciding what goes in its place, and that depends on what the document is for. The three modes can be combined per data type in the same document.

Masking

A solid redaction block or asterisks, preserving the original length so the layout does not shift. For publishing.

Tokenisation

Stable numbered labels: [PER_1] is the same person on page 1 and on page 40. You can tell who is who without knowing who they are. For processing and analytics.

Synthetic data

False but plausible values, of the same type and length, with a valid check digit and consistent gender. For training, testing and statistical transfers.

See the three modes in detail →

Anonymising and pseudonymising are not the same

Two modes

Anonymising is irreversible: no mapping table, no key and no original value is kept. Pseudonymising is reversible: pseudonyms are stable and the mapping is stored encrypted with your organisation key, with reversal restricted to an authorised profile and written to the log.

The difference matters because a pseudonymised document remains personal data under the GDPR, even if whoever reads it cannot identify anyone. It cannot be published as if it were anonymous. anonimiza.do makes you choose the mode, and can produce both outputs in a single run.

Full comparison of the two modes →

Review before you publish

Supervision

Review is carried out by users in your organisation, with their own credentials. The screen shows the document with detections highlighted and, alongside it, the full list: value, data type, confidence and originating layer. One click accepts or discards a detection, and you can manually select a fragment the system did not flag. The «by unique value» mode groups every occurrence of the same value, so a single decision resolves the forty times a name appears in a long case file. Approving generates a numbered version, and earlier ones are kept.

How review works →

Need the technical detail?

The seven layers of the engine, the full data type catalogue with its validations, the format table, the API and the security measures, all on one page.

View technical specifications

Protect the personal data in your documents

3 free documents every month. No credit card. No installation.

Create free account