BONSAI text is a cutting-edge solution designed to securely redact sensitive and personally identifiable information from free-text documents. Our comprehensive application provides a streamlined workflow for detecting and masking confidential content in unstructured text, enabling organizations to use, share, and process documents safely without compromising privacy, security, or control.
By leveraging BONSAI text, organizations can efficiently redact free-text files and generate high-quality outputs for secure document sharing, AI-assisted workflows, and other downstream use cases. The solution combines precise detection of sensitive information with flexible customer-specific exception rules, making it possible to adapt redaction settings to specialized vocabularies, contexts, and operational requirements.
The versatility of BONSAI text extends to supporting multiple text-based file formats, including pdf, txt, and docx documents, while allowing each redaction task to include several files at once. Deployed securely within the customer’s own environment, BONSAI text operates without generative AI or large language models, ensuring that sensitive content never leaves the customer-controlled environment during processing.
BONSAI text workflow has three key phases
Select the text documents and upload them to create a redaction case. Give a name for your case and select how long the case should be stored in the application.
Preview the default redaction of your case documents. Adjust redactions settings by selecting different redaction methods, adding words and terms to case-specific blacklist and whitelist to force them to be either always hidden or visible. You can always choose which masking style to use.
When you’re happy with the redaction settings and how the redacted preview looks, you can download all the redacted files in your redaction case as one zip-file.
Offered currently with fine-tuned Finnish language specific capabilities. Fine-tuning to your local language for accurate redaction can be prepared.
Supports customer-specific exceptions and rules so redaction can reflect organizational policies and specialized vocabulary.
Operates without generative AI or large language models, avoiding the uncertainty of external model-based processing.
Automates a time-consuming compliance task, helping teams move faster while reducing repetitive review effort.
Keep document processing inside your own environment to meet strict security, privacy, and governance requirements.
Automatically detects and masks personally identifiable and confidential information in free-text documents with high accuracy.
Supports fine-tuned redaction for Finnish-language documents as well as separate redaction capabilities for documents in other languages.
Allows a single redaction task to include multiple files, making document handling faster and more efficient.
Processes common text-based file types, including PDF, TXT, and DOCX documents.
Redacted content can be hidden either with black bars or replaced with entity-type labels, depending on workflow needs.
Runs entirely within the customer’s own environment, ensuring that sensitive content remains under customer control.
Operates without generative AI or large language models, so data is not shared with external services during processing.
Customer-specific blacklists and whitelists can be configured to improve precision and reflect domain-specific terminology.
BONSAI text operates on free-text documents as both input and output, supporting widely used file formats such as pdf, txt, and docx. The solution enables intuitive document handling by allowing users to upload one or several files, automatically redact sensitive content, and download the processed outputs keeping the same document look and feel
Efficient multi-file redaction workflows with flexible redaction and masking options for different use cases. Process several documents in a single task and apply case-specific blacklists and whitelists to improve accuracy and reflect domain-specific terminology.
VEIL.AI’s architecture is designed to be robust, secure, and highly adaptable. Whether you are working in a cloud environment, on-premises, or across multiple locations, our technology is engineered to meet your needs.
Available as an Azure Managed Application, BONSAI text makes deployment simple while keeping document processing fully within the customer’s own Azure environment. This gives organizations stronger control over sensitive data, supports compliance and security requirements, and reduces the effort needed to introduce the solution into existing enterprise workflows.
BONSAI text is built for organizations that need to handle sensitive free-text documents with a high level of security and control. Running entirely within the customer’s own environment, the solution ensures that document contents are not sent to external AI providers or third-party processing services.
BONSAI text does not use generative AI or large language models, helping organizations avoid GenAI-related data exposure risks and compliance uncertainties. This makes it easier to align document redaction workflows with internal governance, privacy, and security requirements while enabling safer use of documents in downstream processes.
Our partnerships with major technology companies like Snowflake and Microsoft enhance the functionality and reach of our technology, ensuring that you have access to the best tools and resources available
Our collaborations with leading technology providers ensure that VEIL.AI applications can be easily integrated into your existing data ecosystem. This compatibility enhances your data science capabilities and ensures that your workflows remain uninterrupted.
Dive deeper into VEIL.AI’s technology through our comprehensive whitepapers, case studies, and research articles. Learn how our technology is being applied in real-world scenarios and see the results for yourself.
BONSAI text, for accurate free text redaction We have identified a growing demand among our clients for free text
The recent Court of Justice of the European Union (CJEU) decision in European Data Protection Supervisor (EDPS) v. Single
In 2020, the COVID-19 pandemic clearly revealed that viral diseases recognize no political or geographical boundaries. It also underscored
