PII Redaction

Understand how Hyper handles PII redaction within RAG queries.

PII Redaction

Hyper is committed to ensuring the privacy and security of sensitive data processed through its API, particularly Personally Identifiable Information (PII). Our PII redaction feature is designed to detect and redact PII from both input and output data involved in Retrieval-Augmented Generation (RAG) queries.

How PII Redaction Works

PII redaction in Hyper uses advanced algorithms to identify potential PII in text data. When PII is detected, it is redacted or anonymized before the data is processed or returned by the API. This feature is crucial when dealing with sensitive data in sectors like healthcare, finance, and legal industries.

Input Data Redaction

Before data is processed by Hyper's RAG queries, the PII redaction system scans and redacts any detected PII from the input. This ensures that no sensitive information is stored or used in subsequent processing.

Output Data Redaction

When Hyper generates responses to RAG queries, the output is also scanned for potential PII. Any detected PII is redacted from the output before it is returned to the user, maintaining the privacy of individuals' information.

Configuring PII Redaction

PII redaction can be configured via the Hyper API through specific endpoints. Developers can set their preferred level of redaction and specify which types of PII to target, such as names, addresses, social security numbers, etc.

Use Cases

  • Customer Support: Redact PII from customer queries and the corresponding AI-generated responses to protect customer privacy.
  • Content Generation: Ensure that generated content does not inadvertently include PII, maintaining confidentiality in content such as reports or articles.
  • Data Analysis: Analyze user-generated content without exposing sensitive PII, adhering to data protection regulations.

Best Practices

  • Regularly Update Configurations: As PII detection algorithms evolve, regularly update your redaction configurations to maintain optimal privacy protection.
  • Review Redacted Outputs: Periodically review redacted outputs to ensure that PII is being appropriately handled and that the redaction does not impede the utility of the data.
  • Understand Regulatory Requirements: Familiarize yourself with the data protection laws relevant to your industry and region to ensure your use of PII redaction complies with all regulations.

PII redaction is a dynamic tool within the Hyper API that helps developers handle sensitive data responsibly. By employing PII redaction, you can leverage the power of RAG queries while safeguarding individual privacy and complying with data protection standards.

For more information on enabling and configuring PII redaction in your Hyper integration, refer to the API documentation.