Loading Rampart model…
14.7 MB · ONNX Runtime Web · WASM
Runs 100% in your browser · No data leaves your device

Rampart

A 14.7 MB ONNX model that detects and redacts personally identifiable information before it ever leaves your device — powered by transformers.js running entirely in your browser.

14.7MB
Model Size
17
Entity Types
7
Languages
98.4%
Recall
~4ms
p50 Latency
⚡ Live Redaction
Type or paste text with PII — watch it get redacted in real time
Initializing…
🔒 Redacted output
Start typing above to see PII get redacted…
Rampart detects 17 entity types using a 35-label BIO schema. A deterministic recognizer layer handles structured identifiers (SSN, credit cards, emails, URLs, IPs); the neural model handles contextual PII (names, phones, addresses, document numbers). City, state, and ZIP are retained by default for context.
See how Rampart protects you before sending text to an AI assistant. Your message is redacted client-side, the "assistant" sees only placeholders, and the response is rehydrated with your real values — all in your browser.
Hello! I'm a simulated AI assistant. Type a message below — Rampart will redact your PII before it reaches me, and I'll respond using the placeholders.
Rampart replaces PII with stable, typed placeholders like [GIVEN_NAME_1]. The same value always maps to the same placeholder, so multi-turn conversations stay coherent. The real values live only in the client-side entity table — they never leave your device. Type in the demo above, then check here to see the mapping.
Placeholder → Real Value
Type text in the demo above to see placeholder mappings…