AIGuard
AIGuard
Stop sensitive PII from leaking to external LLMs, and block aggressive AI crawlers from scraping your proprietary content. One unified API proxy and perimeter shield.
Sanitize Prompts Before They Hit External AI Models. We replace sensitive emails, SSNs, and identity data with deterministic tokens before routing to your LLM provider.
Take Back Control of Your Infrastructure. Identify stealth scrapers, calculate dynamic 0-100 risk scores, and generate instant blocking policies for Cloudflare and Nginx.
Automatically captures sensitive data types (emails, SSNs, phone numbers) before they ever leave your server.
Replaces raw data with tokens ([EMAIL_001]) so the LLM responds accurately without seeing real identity data.
Calculates inbound threat vectors (0-100) based on behavior and IP reputation to make crawler threats visible.
Keep using your standard SDKs. No re-writing pipelines. Just change your API base URL and keep building safely.
Point your existing LLM client to `api.aiguard.solutions` instead of the raw provider URL.
Toggle full or partial PII masking and define crawler blocking policies in your AIGuard dashboard.
Deliver safe responses to your users via SSE streaming, while generating transparent compliance logs via Celery workers.
Hard-coded quota limits protect you from Denial of Wallet attacks at every tier.
Your privacy is our architecture. PII tokenization happens entirely in volatile memory. We utilize secure JWT tokens and encrypted API keys, and your raw prompt data is never stored on disk.