PII / NER Redaction
Browser-side sensitive information detection and redaction with deterministic rules and GLiNER ONNX/WebNN enhancement
Category
How to Use
- Paste the text you need to inspect
- Use rule-based redaction or enable GLiNER ONNX/WebNN enhancement
- Review detected emails, phone numbers, names, addresses, and other entities
- Copy the redacted result and manually review high-risk output
Examples
-
Redact contact details
Input:
Alice email is alice@example.com and phone is 13800138000Output:
[PERSON] email is [EMAIL] and phone is [PHONE] -
Rules first
Input:
Order note includes an ID number and phone numberOutput:
Rules detect high-confidence fixed formats before model-assisted entity detection
FAQ
- Is text uploaded?
- No. Text is processed in the browser; model mode also prioritizes local ONNX/WebNN inference.
- Can I use the output as a compliance guarantee?
- No. Treat it as an assistant. Manually sample high-risk data and confirm the required handling for your jurisdiction.
- What is the difference between rules and the model?
- Rules work well for fixed patterns such as email, phone, and ID numbers. The model is better for context-dependent names, organizations, and places.
Related tools
- Word Counter
Count words, characters, lines, and more
- Case Converter
Convert English text to various cases
- Text Diff
Compare two texts and highlight differences
- Markdown Preview
Live preview of Markdown rendering
- Excel to SQL IN
Convert Excel column data to SQL IN statement format with custom delimiters
- 3D Font Generator
Convert text to 3D model with GLB and GLTF format download