Automate Document Redaction

Proactively protect sensitive data with automated document redaction. Leverage AI to redact sensitive information from any document quickly and accurately with guaranteed results.

Schedule Product Demo →

REDACT IMAGES

REDACT TEXT

Powerful Document Redaction with Unparalleled Accuracy

Request Demo

Guaranteed Data Quality

Leverage always the best AI models depending on the task to ensure near 100% detection accuracy. Human-in-the-loop is enabled for post-processing of false positives and negatives for guaranteed quality.

Keep Documents Usable with Pseudonymization

Do not delete data but leverage pseudonymization so that you can process personal data beyond the original purposes and still comprehend contents/contexts and identify patterns and coherences.

Flexible & Efficient

Quickly and easily search and remove relevant information, and apply redaction across an entire page, a range of pages, a full document, or a collection of documents.

Machine Learning-based Redaction

Document.Redact’s detection models are based on machine learning, not rulesets, making them more robust to non-standard patterns, regional specifics and custom requirements.

Redaction That Scales

Near infinite scalability, low latency, and 24/7/365 availability means Document.Redact scales up (or down) instantaneously as your needs change.

Any Document Type

Document.Redact is suitable for any type of pdf and Microsoft Word document, incl. embedded data, regardless of source, format, layout or structure.

How does Document.Redact work?

Choose from a list what you want to redact. These settings can be changed anytime. In case you feel the list is incomplete, we’re happy to add choices for you.

By choosing a setup of workers to suit your unique needs, you’re in full control over costs, speed and quality.

Super.AI’s platform routes the redaction task to multiple machine learning-based NLP models to understand the text written in the document(s).

The results of multiple NLP-algorithms are automatically compared, and combined if necessary, to return the most accurate anonymization result.

Users can easily and quickly review redaction results and, if necessary, add, delete or manipulate redaction. In the given example, the blur around the face has been enlarged.

Document post-processing engines obfuscates the entities/areas of interest. Machine-readable PDFs get redacted on a text level (each character is replaced with an X).

Document post-processing engines obfuscates the entities/areas of interest. Scanned documents and embedded images are redacted on the pixel level (entities are blurred using a Gaussian filter).

Super.AI Document.Redact Overview

View and download a PDF overview of super.AI’s Document.Redact solution

Document Redaction for
Every Use Case

From redacting inspection dates to obfuscating faces in embedded images, we’ve got it covered.

Text Redaction

Automatically detect and redact text, including names (people, organizations, etc.), phone numbers, dates, street addresses, email addresses, vehicle identification numbers and more. Document.Redact can recognize and remove handwritten text and understand multiple languages.

Learn more

Image Redaction

Automatically detect and redact embedded graphics and images. Personally identifiable information like faces and license plates can be automatically identified and removed from any document. Document.Redact leverages multiple AI models to detect and redact virtually any type of unstructured image data.

Learn more

Document Redaction for Every Industry

Banking & Finance

Protect sensitive data and customer information at scale. Automatically redact account numbers, social security numbers, credit card numbers, and other sensitive information.

Logistics

Automatically redact sensitive information from arrival and delivery notices to enable better record-keeping without while mitigating risk. Make data usable for internal training and documentation.

Legal

Automatically redact privileged and confidential information from business records, financial statements, and any other document types while complying with privacy regulations like GDPR.

Testing, Inspection, & Certification (TIC) Services

Anonymize safety assessment audits and certifications, automotive inspection reports, vehicle condition inspection reports and damage and value assessment reports.

Additional Redaction Resources

Blog

Frequently Asked
Questions

Technical and Product

Data Privacy & Security

Billing & Payment

Technical & Product

What kind of information can be redacted?

Currently, users can redact person names, organization names, street addresses, dates, phone numbers, report numbers, email addresses, license plates, and vehicle identification numbers in texts in documents. If you're interested in more, e.g. embedded images or graphics, please contact us to inquire about a customized version.

Can I use your software for any language, document type, etc. or do you need to train the models first?

Document.Redact is agnostic and works out of the box. Currently, we support English and German. If you need help processing documents in a different language, please contact us to discuss how we can accommodate your request.

Can I alter the output?

Yes, you can use the graphical user interface for the correction of false positives, meaning adding and/or extending bounding boxes for specific object classes, and/or correction of false negatives, meaning de-selecting blurred objects.

Can the redaction be reverse engineered?

No. We erase information from the document and replace it with pseudonyms (e.g. “Person A” and “Person A Street Adress”). If you’re interested in document anonymization, contact us to discuss how we can accommodate your custom request.

What’s the detection accuracy of your solution?

Results of different detection models are compared and the best result is used. This leads to higher accuracies compared to using only one single model, so we’re confident to guarantee 99+% detection accuracy.

What is Document.Redact?

Document.Redact is an auto-redaction service that automatically removes personal and sensitive information from documents.

What data types are supported?

Currently, users can redact names (people, organizations, etc.), street addresses, dates, phone numbers, report numbers, email addresses, license plates, and vehicle identification numbers in texts and documents. If you're interested in more (e.g. embedded images or graphics) please contact us to inquire about customization.

Can I integrate the product via API?

Yes. Please see our API reference, which covers every endpoint and parameter. This enables easy integration of super.AI products into existing applications or services.

Does the product offer a graphical user interface (GUI)?

Yes. Besides the API, users can access the service via a graphical user interface (GUI) that requires no coding experience.

Does the service have minimum hardware or software requirements?

No. When using the service through our API, all that is required is a stable internet connection. Simply create an account and start using the platform immediately.

Do your AI solutions work in any scenario or is training required to work with new datasets?

Our AI-based service is a generally applicable, out-of-the-box solution. However, we offer custom model training to accommodate unusual settings and/or perspectives. Please contact us for more information.

Data Privacy & Security

Do I have control over how my data is stored?

Yes. Customers decide if project data is stored or deleted. After a contract termination, all customer data is deleted within 90 days unless specified otherwise. For more information, please see our data privacy and security documentation. For specific questions, please email: security@super.ai.

Which security standards and compliance certifications do you support?

We support SOC, SOC II, HIPAA/HITECH, 2020 Act, APPI, APPs, CSL, CCPA, CPPA, GDPR, LGPD, PDPL, PIPA, PIPL, POPIA, UK-GDPR, as well as ISO/IEC 27001, 27017, 27018, 9001, FIPS 140-2, NIST 800-171 and CSA STAR CCM v3.0.1.

Do your products comply with applicable data protection laws, like the GDPR or CCPA?

Although no single product can ensure compliance with GDPR or CCPA on its own, super.AI's data privacy and security practices make it east to keep your data in compliance with applicable laws. For specific questions, please email: security@super.ai.

Where is the service hosted?

The super.AI platform is built on top of Amazon Web Services (AWS) and is hosted in the US-East (Virginia) AWS Region. Please contact us to inquire about deployment in another region.

Billing & Payment

Is there a free version?

Yes. Our 7-day free trial offers full product functionality. No credit card is required for signup, simply make an account and start processing unstructured data with AI. After the trial period ends, a plan must be purchased to continue using the platform or via contacting us.

What payment methods are available?

We accept credit cards for the standard plans. If you require a different method, please contact our sales team.

Automate Document Redaction

Powerful Document Redaction with Unparalleled Accuracy

How does Document.Redact work?

Define what to redact

Full Controllability

AI-automated detection

Fully Automated Quality Assurance

Audit/Markup Copy for Manual QA

Non-reversible anonymization - Option 1