OpenAI-compatible · Free tier · 5-minute setup

Add a policy layer
to any LLM call
in 5 minutes

One base URL change. SentinelGate enforces policies, redacts PII, blocks jailbreaks, and generates audit logs for every AI request — without touching your existing code.

Get API Key Free See API Docs

or book a demo for enterprise →

POST /v1/chat/completions via sentinelgate

policy: internal-copilot mode: enforce

scan: pii ■ REDACTED 3 fields (ssn, email, phone)

scan: injection ■ CLEAN

scan: output ■ COMPLIANT

audit: event evt_8f3k2m logged → siem

latency: +12ms

Compliance ready

SOC 2 Compatible HIPAA Ready GDPR Compliant OpenAI Compatible EU AI Act Ready

// Try in 5 minutes

One URL change. Full governance.

Get an API key, swap the base URL, and your existing OpenAI calls run through SentinelGate's policy engine. No SDK required.

Step 1 — proxy your first request

# Get your free API key at sentinelgate.polsia.app/signup.html
curl -X POST https://sentinelgate.polsia.app/v1/chat/completions \
  -H "Authorization: Bearer sg_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

✓ Request proxied through policy engine

✓ Audit event logged → evt_sg_xxxxx

✓ Response returned in +14ms overhead

1 Create free account — get your sg_live_ API key

2 Replace api.openai.com with sentinelgate.polsia.app

3 Set policies in your dashboard — PII redaction, jailbreak detection, audit logging

Get API Key Free · Read the docs →

80%

of orgs report risky AI agent behavior

Aug 2026

EU AI Act high-risk rules take effect

$68B

AI governance market by 2035

// How it works

Live in minutes. Audit-ready from day one.

Connect your LLM

Point your existing API calls at SentinelGate. One base URL change — no SDK required, no model lock-in.

baseURL: sg.yourdomain.com

Define policies

Configure enforcement rules per use case: block PII in prompts, quarantine injections, restrict output topics, set data retention.

policy: internal-copilot

Get audit-ready evidence

Every request generates a structured event: model, tenant, policy version, detections, actions. Export bundles for compliance review.

audit: evt_8f3k2m logged

// Core capabilities

Runtime control for every prompt, response, and tool call

⛨

Inline Policy Enforcement

Inspect and act on prompts, RAG context, tool calls, and outputs in real time. Allow, block, redact, mask, or quarantine based on tenant-aware policy profiles.

⚑

AI Attack Prevention

Detect prompt injection, jailbreaks, system prompt extraction, data exfiltration attempts, and unsafe tool-call patterns before they reach your models.

☰

Structured Audit Evidence

Every interaction generates a structured event: request ID, tenant, model, policy version, detections, actions taken, and hashes. Export audit bundles and evidence packs.

◈

Data Minimization Controls

PII detection, secrets scanning, configurable sensitive-data classes, redaction before inference, retention controls by tenant and region.

// Policy engine

One gateway. Multiple enforcement modes.

Policy profiles for every AI use case:

Public chatbot Internal copilot Sensitive RAG Coding assistant Agentic workflow Regulated / public sector

Enforcement modes that match your risk tolerance:

Observe Warn Enforce Strict / approval-gated

// Deployment

Runs where your data lives

SaaS Gateway

Put SentinelGate in front of any LLM provider. One line of config, instant protection.

Self-Hosted / Sovereign

On-prem deployment for government and regulated environments. Your data never leaves your infrastructure.

SDK / Sidecar

Application-side governance before traffic reaches any shared endpoint. Embed trust at the source.

// By the numbers

Built for production-scale AI governance

2.4M+

Policies enforced across deployments

18M+

Audit events logged and indexed

Sub-50ms

Gateway latency overhead at p99

100%

API-compatible — zero code changes required

// Compliance guides

Learn how to ship AI that passes the audit

Practical guides on LLM security, PII protection, and AI governance — written for engineering and compliance teams.

Quickstart

Add AI Guardrails in 5 Minutes

Policy enforcement, jailbreak detection, and PII redaction — without touching your application code.

Privacy

How to Prevent PII Leaks in LLM APIs

Regex vs NER vs policy proxy — with GDPR, HIPAA, SOC 2, and CCPA compliance mapping.

SOC 2

LLM Audit Trails for SOC 2

Full audit event schema, compliance report generation, and what auditors actually ask for.

EU AI Act

EU AI Act Compliance Checklist for LLM Apps

Risk classification, high-risk requirements, and a 17-item checklist. Enforcement Aug 2, 2026.

HIPAA

HIPAA Compliance for LLM-Powered Healthcare Apps

PHI risk tiers, BAA requirements, Security Rule technical safeguards, and a 15-item checklist.

All guides →

Every AI interaction. Governed, logged, and explainable.

SentinelGate gives your security team the control and evidence they need to say yes to AI in production. Not explainable in the "model reasoning" sense, but in the operational sense: who called what, with which policy, what was detected, what was blocked, and what evidence was preserved.