The Challenge
Development teams face a constant tension between realistic test data and data protection:
- Production data contains sensitive PII that cannot be used directly
- Synthetic data often lacks the realism needed for meaningful testing
- Manual anonymization is time-consuming and error-prone
- Different environments need consistent, reproducible data
The Solution
Integrate PII anonymization directly into your development workflow with our RESTful API.
REST API
Simple JSON API for integration into any stack. Analyze and anonymize in a single request.
CI/CD Ready
Automate test data generation in your pipeline. Consistent results every time.
Reproducible
Same input, same output. Deterministic results for reliable testing.
Fast
Process thousands of records per minute. No GPU required.
Frequently Asked Questions
How do I integrate cloak.business PII detection into my CI/CD pipeline?
Use the REST API or official SDKs (JavaScript/Python) to call the analyze and anonymize endpoints from any CI/CD tool. The API returns deterministic results, so the same input always produces the same tokens — ideal for snapshot testing and reproducible builds.
Can cloak.business generate realistic anonymized test data from production data?
Yes. By anonymizing production exports with the Replace or Encrypt method, you get structurally realistic test data with no real PII. The Encrypt method is reversible, so you can restore originals for debugging when authorized.
What programming languages does the cloak.business SDK support?
Official SDKs are available for JavaScript (npm: @cloak-business/sdk) and Python (PyPI: cloak-business). Both SDKs auto-retry on HTTP 429 with Retry-After and support all three API endpoints: analyze, anonymize, and deanonymize.
Is This Right for You?
Best For
- Organizations with compliance obligations (GDPR, HIPAA, CCPA, PCI-DSS)
- Teams regularly sharing datasets containing names, IDs, or medical records
- Developers building AI pipelines that process user-submitted content
- Enterprises requiring audit logs and reproducible anonymization for legal holds
Not Ideal For
- Single-language English-only pipelines with no PII — regex-only tools may suffice
- Real-time streaming at sub-5ms latency — NLP inference adds overhead
- Fully air-gapped environments without internet access — use Desktop App instead
- Unstructured media files (audio, video) — text extraction is a prerequisite limitation