Stop sensitive data
before it leaves
your perimeter.
PromptGate intercepts every LLM request — scanning prompts, responses, and file attachments for GDPR, HIPAA, PCI-DSS, and PII violations. Redact, block, or alert in real time.
Proxy. Scan. Redact. Ship.
PromptGate sits inline between your app and any LLM provider — zero code changes required.
Route your LLM traffic through our proxy endpoint. Supports OpenAI, Anthropic, Google, Azure, and any OpenAI-compatible API.
Prompt text, system messages, and attached files are inspected against 40+ regulatory patterns including GDPR, HIPAA, PCI-DSS, and custom rules.
Configure policy actions per finding type: hard block, auto-redact sensitive spans, mask file content, or emit a compliance alert to your SIEM.
The sanitized request is forwarded to the LLM. Your application receives the response normally, with an audit trail for every scan.
Sensitive data, surgically removed.
PromptGate identifies and redacts PII, PHI, and financial data — in both free text and structured files — before a single token reaches the model.
admitted 14 March 2024. Primary contact email:
GDPR. Phone: PII.
Payment on file: PCI-DSS.
Request originated from IP.
Enterprise-grade guardrails,
developer-friendly setup.
Everything you need to deploy safe AI workflows — from single startups to regulated enterprises.
Every user message and system prompt is scanned before reaching the LLM. Regex, NER, and ML-based detectors catch what rules alone miss.
PDFs, DOCX, XLSX, images, and CSVs are parsed and deep-scanned. Sensitive content is redacted in the extracted text before ingestion.
LLM outputs are scanned on the way back. If the model inadvertently leaks or reconstructs sensitive data, PromptGate catches it before your app sees it.
Define rules in YAML or via the dashboard. Set actions per violation type: redact, block, warn, or log. Version-control your policies alongside your codebase.
Every scan result is logged with timestamps, user context, and matched patterns. Export to your SIEM or pull via the audit API for compliance reporting.
Train PromptGate on your own data vocabulary. Proprietary contract IDs, internal project codes, trade secrets — define what's sensitive for your business.
Built for regulated industries.
PromptGate ships with pre-built detection patterns for the frameworks that matter most.
Simple, usage-based pricing.
Start free. Scale as your LLM usage grows. No per-seat nonsense.
For developers exploring safe LLM integrations.
- 500K tokens scanned / month
- Core HIPAA + GDPR detectors
- Prompt + response scanning
- API proxy access
- 7-day audit log retention
For teams building production AI features.
- 10M tokens scanned / month
- All 40+ compliance frameworks
- File attachment scanning
- Auto-redaction engine
- Webhook + SIEM integration
- 90-day audit log retention
- Custom entity detection
For regulated industries with strict compliance requirements.
- Unlimited token scanning
- On-premise deployment
- VPC / private cloud
- SOC 2 Type II reports
- Dedicated SLA
- Custom retention + DLP
- SSO / SAML / SCIM
Your LLM pipeline
deserves a gatekeeper.
Start scanning in minutes. No infrastructure changes. Just point your API calls through PromptGate.