Platform
Platform overview Six modules from discovery to evidence on one governance backbone. AI Inventory & Discovery Make shadow AI usage visible. Use-case Intake & Triage One-page intake, 5 business-day SLA. Risk & Autonomy Tiering Green / Amber / Red decision model. AI Gateway & Policy All model traffic through one control point. Audit Trail & Evidence Audit-ready, tamper-evident record.
Solutions
AI Center of Excellence A single operating console for the hub team. CISO & Security Data-leak control and kill switch. Legal & Compliance Collect regulatory evidence automatically. Executive Leadership Measure the return on AI investment. Business Units (Spoke) Make the approved path the easy path. Industries Manufacturing, finance, health and public sector.
Compliance
EU AI Act Obligation timeline and readiness map. ISO/IEC 42001 AI management system (AIMS) controls. NIST AI RMF Govern, Map, Measure, Manage mapping. KVKK & GDPR Personal data and DPIA linkage. Security & Architecture Tenant isolation, SSO, MFA, data residency. Control matrix Which module satisfies which clause.
Framework
Resources
About Contact Customer login Request a demo
Security & Architecture

Tenant data is isolated at the application layer.

An AI governance platform holds an organisation's most sensitive inventory: which team does what, with which data. That is why isolation, identity and auditability are designed in from the start.

Foundations

Three foundational architecture decisions

Tenant isolation, strong identity and tamper-evident records.

Tenant isolation

Every customer is an Organization. Every tenant-scoped table carries an organisation identifier, and isolation is enforced at the application layer.

  • Row-level scope enforcement
  • Cross-tenant queries blocked
  • Automated isolation tests

Identity and access

Connect to your corporate identity provider with per-tenant OIDC configuration; where local login is used, TOTP-based multi-factor authentication is mandatory.

  • Per-organisation OIDC/SSO
  • Mandatory TOTP MFA
  • Short-lived access + refresh tokens

Audit and accountability

Every critical action, including permission changes, is written to the audit trail with its actor and rationale, and can be exported.

  • Actor, action, entity and metadata
  • Role assignment records
  • Export for audit
Multi-tenancy

One tenant's data is never visible to another.

On a multi-tenant platform the biggest risk is a single query that accidentally escapes its scope. That is why isolation is a tested behaviour, not a hoped-for assumption.

  • An organisation identifier is mandatory on every tenant-scoped table.
  • Access scope is constrained at the application layer by the session's organisation.
  • Cross-tenant super-admin authority is separated behind a distinct flag (is_platform_owner).
  • Tenant isolation is continuously verified by the automated test suite.

Role-based access at the data level, not the menu level

A spoke representative sees only their own business unit's inventory, and that constraint is applied through query scope rather than by hiding parts of the interface. A request that bypasses the UI and calls the API directly hits the same boundary.

Controls

Application security controls

The following controls are part of the platform's standard configuration.

Area Control Note
Authentication OIDC/SSO or local password with bcrypt Configured per tenant
Multi-factor authentication Mandatory TOTP MFA Required for local password login
Session Short-lived access token + refresh Role claims mapped at token issuance
Authorisation Role-based access control (RBAC) Scope enforced at the data layer
Transport security HTTPS/TLS Terminated at the reverse proxy
Application headers Security headers and CORS policy helmet-based configuration
Abuse prevention Rate limiting Tightened on authentication endpoints
Input validation Schema-based request validation On all write endpoints
Auditability Central audit trail and structured logging Exportable
Responsible disclosure

If you found a vulnerability, we want to hear about it.

We take reports from security researchers seriously. Please send your finding to our security team together with reproduction steps. We will confirm receipt and keep you informed as the process progresses.

  • Give us a reasonable remediation window before disclosing publicly.
  • Do not access customer data, modify data or disrupt the service.
  • Test only against your own tenant and your own account.
  • State reproduction steps and impact clearly in your report.

Let us run a technical session with your security team.

Discuss architecture, data residency and deployment model questions directly with our engineering team.