AUTOMATED AI TRiSM & COMPLIANCE

Secure the Sovereign AI
Frontier

Deterministic 100% air-gapped compliance verification and automated technical evidence mapping for ISO/IEC 42001 & the EU AI Act.

github actions — neural-sentinel-scan.yml
$Running Neural Sentinel v2.0...
✅ Valid ENTERPRISE license verified cryptographically.
→ Scanning agent configuration for structural vulnerabilities...
[HIGH] CWE-269 — Unconstrained parameter detected in 'execute_tool' schema
→ Mapping to ISO/IEC 27001 A.8.12 & ISO 42001 A.7.2...
[COMPLIANCE] Missing required risk mitigation control.
→ Action Failed: Blocking PR merge. Exit code 1.
$

<60sScan Time
100%Air-Gapped
ISO42001 Mapped
EUAI Act Ready

// live demo

See a Scan from Start to Finish

Watch Neural Sentinel scan an AI agent configuration, enforce cryptographic licensing, and map each finding to compliance evidence.

Disclaimer: This video is for educational and demonstration purposes only. It showcases Neural Sentinel’s automated detection capabilities on publicly available, open-source repositories. This does not constitute an official security audit, nor does it imply any affiliation, sponsorship, or endorsement by Microsoft or any other third-party organization.


// scan output

Every scan generates five files

One scan produces the right deliverable for each person who needs it, from the board to the build pipeline.

_FINAL.pdf
CISOs & Boards
Executive summary of risk and compliance posture (Dynamic White-Label available).
_report.md
Developers
Actionable remediation steps and exact line numbers for each finding.
_audit_matrix.md
Auditors
Direct control mapping to ISO/IEC 27001:2022, ISO/IEC 42001, and EU AI Act.
_merged.json
GRC platforms & SIEMs
Machine-readable, HMAC-SHA256 signed data for automated GRC ingestion (Webhook delivery unlocked in Enterprise/GRC tiers).
_manifest.json
Proof of work
The AI tool schemas extracted during the scan. Never leaves your environment.

All five files are written inside your own environment. _manifest.json never leaves it.
Note: Neural Sentinel automates technical evidence gathering for specific controls. It supports, but does not replace, a formal certification audit.


// architecture

Built for Developers.
Trusted by Auditors.

We replace manual AI compliance verification with deterministic , automated evidence collection, without compromising data privacy.

⚡

Developer GitHub Action

A lightweight, shift-left SAST tool that catches structural AI vulnerabilities (e.g., unconstrained parameters, prompt injection risks) and blocks PRs automatically. Free for open-source.

🔒

Enterprise Docker Appliance

A self-hosted, air-gapped container that performs deterministic AST scanning of AI agent configurations. Generates structured JSON compliance matrices and auditor-ready PDF reports. Zero data leaves the client's VPC.

⚖️

Automated GRC Mapping

Automatically translates AI-specific technical vulnerabilities into evidence mapped to critical ISO/IEC 27001:2022 (e.g., A.8.8, A.8.12) and ISO/IEC 42001 controls, ready for direct API ingestion into major GRC platforms.


// pricing & partners

Scale Your AI Compliance Offerings

We do not compete with your services; we empower them. Our transparent "Platform + Client Block" model ensures you only pay for the value you extract, with unlimited air-gapped scans included.

Solo / Internal
$1,490
/ year
  • For internal security teams & solo consultants
  • Unlimited air-gapped scans for internal use only
  • 5-file output per scan + SARIF export
  • Direct founder support & onboarding
  • Standard Neural Sentinel branding on reports
Request 35-Day NFR Evaluation →
Enterprise MSSP
$14,900
/ year
  • For established firms managing large portfolios
  • Includes up to 15 Registered End-Clients
  • Multi-tenant GRC Webhook Routing (Vanta, Drata, etc.)
  • Bulk cryptographic key generation dashboard
  • Priority SLA & dedicated technical account manager
  • + $1,500/year per additional registered client (Volume Discount)
Contact Sales →
💼 Need Board-Level Technical Validation?
For high-stakes enterprise or government submissions, partners can add a Neural Sentinel Technical Validation Review:
• Automated Technical Readiness Report ($5,000 / engagement): A premium, vendor-signed summary of scan results mapped to ISO/IEC 42001 & EU AI Act.
• Expert-Led Architecture Review ($15,000 / engagement): Includes the Report plus a 1-hour technical walkthrough with a Neural Sentinel engineer to defend findings to auditors.
*Note: Neural Sentinel provides automated technical evidence and readiness validation. It supports, but does not replace, formal certification by an accredited third-party conformity assessment body.
💼 Not sure which tier is right? We provide 35-day Not-For-Resale (NFR) evaluation keys for internal lab testing and technical validation, no strings attached.
Request 35-Day Access →

// get started

Request 35-Day Evaluation Access

Tell us about your team and what you want to validate. We manually review requests and reply with a 35-day NFR license key and next steps.

Use your company email address, not a personal one (gmail.com, yahoo.com, etc.)


// faq

Common questions

Does Neural Sentinel send our code or prompts to the cloud?

Absolutely not. Our core differentiator is the Enterprise Docker Appliance, which is 100% air-gapped. It runs locally within the client's own VPC or on-premise infrastructure. Zero data ever leaves your environment.

How is this different from standard SAST tools?

Standard SAST looks for generic code flaws. Neural Sentinel is a deterministic compliance verifier purpose-built for AI architectures. It detects AI-specific structural vulnerabilities (e.g., excessive agency, unconstrained tool execution) and maps them directly to technical controls in frameworks like ISO/IEC 42001 and the EU AI Act.

Can Neural Sentinel automatically push findings to our GRC platform?

Yes. Our Boutique Agency and Enterprise MSSP tiers include automated, HMAC-SHA256 signed webhooks that push the _merged.json compliance evidence directly to platforms like Vanta, Drata, or Jira. Only the sanitized, signed JSON payload is transmitted.

How does your partnership licensing work?

We use a transparent "Platform + Client Block" model. You purchase a base tier (e.g., Boutique Agency at $4,900/yr) which includes the white-label Verifier and your first 3 Registered End-Clients. As you win new business, you simply add clients at $2,500/year each. We conduct an annual usage reconciliation to ensure partners are properly licensed for their scale.

How do we start a partnership or evaluation?

Step 1: Contact us using your work email to receive a 35-day NFR (Not-For-Resale) evaluation key. This gives your technical team full access to validate the Docker appliance in your lab.
Step 2: Once validated, we discuss the best fit for your volume, whether that’s our Boutique Agency tier or our Enterprise MSSP tier, and get your first Registered End-Clients set up.