GitHub
AI Security Proxy

Proxilion GRC

Stop Sensitive Data from Leaking to AI

Security proxy that intercepts all traffic to ChatGPT, Claude, Gemini, and other AI services. Scan for PII, enforce compliance policies, and maintain complete audit trails.

30+
PII Patterns
23+
Compliance Standards
6
AI Providers
MIT
Licensed

How It Works

Proxilion GRC sits between your users and AI providers, scanning every request before it leaves your network.

User RequestProxilion GRCAI Provider

|
[Security Scanning]
PII Detection • Compliance Rules • Policy Enforcement • Audit Logging

PII Detection

30+ patterns with validation: credit cards (Luhn), SSNs, IBANs, routing numbers, Medicare IDs, driver's licenses, passports, and more.

Policy Engine

Priority-based rules with configurable actions: BLOCK, ALLOW, MODIFY, ALERT, LOG, QUEUE, or REDIRECT. Set threat levels from LOW to CRITICAL.

Compliance Scanning

Pre-built rules for HIPAA, PCI-DSS, SOX, GLBA, CCPA, GDPR, PIPEDA, SOC 2, ISO 27001, NIST, and 12+ more standards.

SIEM Integration

Forward events to Splunk, QRadar, ArcSight, Sentinel, or Elastic. Supports CEF, LEEF, JSON, and Syslog formats.

Rate Limiting

Four algorithms: token bucket, sliding window, fixed window, and leaky bucket. Protect against abuse and control costs.

Streaming Support

Real-time SSE processing with PII redaction as tokens stream. No delay in user experience.

Supported AI Providers

Proxy traffic to any major AI service. One security layer for all your AI tools.

Provider Models Features
OpenAI GPT-4, GPT-3.5, ChatGPT Full request/response scanning, streaming
Anthropic Claude 3, Claude 2 Full request/response scanning, streaming
Google Gemini, PaLM Full request/response scanning, streaming
Cohere Command, Generate Full request/response scanning
AWS Bedrock All Bedrock models Full request/response scanning
Azure OpenAI All Azure-hosted models Full request/response scanning, streaming

Deployment Options

Deploy where it makes sense for your infrastructure.

Cloudflare Workers

Edge deployment at 300+ global locations

Docker

Single container deployment

Kubernetes

Scalable cluster with HPA

Self-Hosted

Direct Node.js on Linux

Use Cases

Why organizations deploy Proxilion GRC.

Prevent Data Leakage

Employees paste sensitive data into ChatGPT without thinking. Credit cards, SSNs, customer data, source code. Proxilion catches it before it leaves your network.

Regulatory Compliance

HIPAA requires you to protect PHI. PCI-DSS requires you to protect cardholder data. Demonstrate to auditors that AI traffic is scanned and controlled.

Cost Control

Track AI usage per user, team, and model. Set budget limits with threshold alerts. Prevent runaway costs from enthusiastic adoption.

Security Monitoring

Forward all AI interactions to your SIEM. Detect anomalous usage patterns. Maintain complete audit trails for incident response.

Observability Built In

Monitor everything. Prometheus metrics, OpenTelemetry tracing, pre-built Grafana dashboards.

Prometheus Metrics

Export metrics in Prometheus format. Track requests, latency, PII detections, policy violations, and more.

OpenTelemetry Tracing

Distributed tracing for every request. Debug latency issues. Understand the full request lifecycle.

Grafana Dashboards

Pre-built dashboards included. Visualize traffic, detections, latency, and compliance status at a glance.

Secure Your AI Traffic Today

Open source. Self-hosted. Full control over your data.

View on GitHub