SECURE YOUR AI. PROTECT YOUR DATA.

> Enterprise security for LLM applications — detect threats, block PII, audit everything

✓ BROWSER EXTENSION ✓ SECURITY PROXY ✓ SELF-HOSTED
START FREE TRIAL VIEW ON GITHUB

// Platform v0.7.0 · Proxy v0.8.1 · Extension v1.0 (pending review) | MIT Licensed | Supports HIPAA, GDPR, PCI compliance workflows

🚀

Now Live: Platform v0.7.0 + Proxy v0.8.1

What's new: Self-service platform with API key management, PII detection, quota tracking, and Sentry monitoring. Proxy v0.8.1 verified on Windows and Linux with <1ms latency.

Chrome Extension v1.0 submitted to Chrome Web Store (expected approval ~March 1, 2026). — View changelog →

// THREE WAYS TO PROTECT YOUR LLM APPLICATIONS

🧩 Browser Extension

Status: Pending Chrome Web Store review (~7 days)
  • Universal Coverage: Works on ChatGPT, Claude, Gemini, any LLM web app
  • Real-Time Alerts: Popup notifications for detected threats
  • Privacy-First: Analysis runs locally, optional cloud sync
  • Zero Config: Install, sign up, browse — it just works
  • Early Access: Email [email protected] for dev build

📡 Security Proxy (v0.8.1)

Status: Production-ready on Windows and Linux
  • Drop-In Replacement: OpenAI-compatible, one-line code change
  • Advanced Detection: Blocks PII, prompt injection, data exfiltration
  • Multi-Provider: OpenAI, Anthropic, Google, Cohere, local models
  • < 1ms Latency: Minimal overhead, production-grade performance
  • Self-Hosted: Your data never leaves your infrastructure

🖥️ Platform (v0.7.0)

Status: Production-ready with Sentry monitoring
  • Self-Service Onboarding: Sign up, generate API keys, start protecting
  • API Key Management: Unlimited keys, environment tagging, usage tracking
  • Usage Quotas: Free tier (100 req/month), track consumption in real-time
  • Demo Mode: Try InferShield without signup (10 requests)
  • Real-Time Dashboard: Monitor threats, view audit logs, track risk scores

// LIVE PII DETECTION

infershield-proxy@localhost:8000 - v0.8.1
[00:12:03] INFO PII detection enabled (15 pattern types)
[00:12:05] INFO Connected to OpenAI API (gpt-4)
[00:12:08] PASS Request ID: req_abc123 | Risk: 12 | PII: 0 | Latency: 0.8ms
[00:12:12] DETECT PII detected in prompt (3 instances)
└─> Type: SSN (123-45-6789) — Action: REDACTED → XXX-XX-6789
└─> Type: Credit Card (4532-1488-0343-6467) — Action: REDACTED → [CC_REDACTED]
└─> Type: Email ([email protected]) — Action: REDACTED → [EMAIL_REDACTED]
[00:12:13] PASS Request ID: req_def456 | Risk: 45 | PII: 3 redacted | Latency: 0.9ms
[00:12:15] DETECT Medical record number + API key detected
└─> Type: MRN (MRN-987654) — Severity: CRITICAL (HIPAA)
└─> Type: API Key (sk-proj-...) — Severity: CRITICAL (Secrets)
└─> Risk Score: 92/100
└─> Action: BLOCKED
[00:12:16] BLOCK Request ID: req_ghi789 | Risk: 92 | PII: 2 critical | Compliance: HIPAA

// HOW IT WORKS

YOUR APP
Sends LLM request with SSN, credit card, etc.
→
INFERSHIELD PROXY
localhost:8000 (v0.8.1)
✓ Detect threats
✓ Block PII/secrets
✓ Log & audit
→
LLM PROVIDER
OpenAI, Anthropic, Google, etc.
1

Install InferShield

Choose your deployment: Browser extension (Chrome), security proxy (Docker/npm), or platform (self-hosted).

docker run -p 8000:8000 infershield/proxy
2

Configure Detection

Set your risk threshold, choose PII redaction strategy, enable compliance mode (HIPAA/GDPR/PCI).

RISK_THRESHOLD=80 BLOCK_HIGH_RISK=true
3

Update Your Code

Change your API endpoint to point to InferShield proxy. One line change, zero refactoring.

base_url="http://localhost:8000/v1"
4

Monitor & Audit

View real-time threat detection in dashboard. Export audit logs for compliance. Block high-risk requests automatically.

Dashboard: localhost:3000

// THE COMPLIANCE CRISIS

HEALTHCARE QUESTION
"How do I let doctors use ChatGPT?"
without violating HIPAA and leaking patient data
FINANCE QUESTION
"How do we enable AI?"
without exposing account numbers and payment cards
LEGAL QUESTION
"How do we use AI?"
without violating attorney-client privilege
⚠ [HIPAA]

Protected Health Information

Medical records, SSNs, dates of birth leaked to OpenAI. $50k+ HIPAA fine per violation.

⚠ [PCI-DSS]

Payment Card Data

Credit card numbers sent to LLM providers. Loss of PCI compliance, audit failure.

⚠ [GDPR]

Personal Data

Email addresses, phone numbers, IP addresses collected without consent. €20M+ fine.

⚠ [SOC 2]

Enterprise Security

No audit trail. No access controls. No data classification. Audit failure imminent.

// PRICING

Free

$0/month
  • ✓ 100 requests/month
  • ✓ Basic PII detection
  • ✓ MASK redaction
  • ✓ Community support
  • ✓ Self-hosted deployment
  • ✓ MIT licensed source code
Get Started
MOST POPULAR

Pro

$99/month
  • ✓ 10,000 requests/month
  • ✓ All 15 PII types
  • ✓ PARTIAL + HASH redaction
  • ✓ Compliance reports
  • ✓ Email support
  • ✓ Prometheus metrics
Start Free Trial

Enterprise

$499/month
  • ✓ Unlimited requests
  • ✓ Everything in Pro
  • ✓ TOKEN + REMOVE strategies
  • ✓ Custom PII patterns
  • ⏳ SSO (SAML, OAuth) (Planned Q2 2026)
  • ✓ SLA + Dedicated support
Contact Sales

// All plans include full access to open-source code. Paid plans add managed cloud hosting, compliance features, and support.

// GET STARTED TODAY

Built for teams shipping AI applications. Start your free trial today — no credit card required.

START FREE TRIAL (100 REQ/MONTH) SELF-HOST (DOCKER/K8S)

// MIT Licensed | Self-hosted or cloud | No vendor lock-in