Security & AI governance

Enterprise security, built in from day one.

Tenant data is isolated at the database layer, every action is logged, and AI never publishes without a person signing off.

Source-grounded answers only
Prompt and generation history logged
Human approval before publishing
No training on client data
Defence in depth

Five layers around your data.

Hover or tap a layer to see what it does.

  1. Identity

    Single sign-on with SAML 2.0, OIDC and OAuth providers, plus MFA with authenticator apps and passkeys. Passwordless sign-in with magic links.

  2. Access

    Platform, tenant, course and group roles, with permission history and password policies.

  3. Isolation

    PostgreSQL Row-Level Security keeps each tenant’s data separate at the database layer.

  4. Audit

    Immutable audit logs record every action, ready for regulators and internal review.

  5. AI governance

    AI outputs are logged, labelled and editable, and nothing publishes without human approval.

What protects your data

Controls at every layer.

Database-level tenant isolation

PostgreSQL Row-Level Security keeps each organisation’s data separate, not just filtered in the app.

Single sign-on

Google, Microsoft Entra, Okta, OneLogin and any SAML 2.0, OIDC or OAuth provider.

Multi-factor authentication

TOTP and WebAuthn, including passkeys, hardware keys and recovery codes.

Immutable audit log

Every action recorded with actor, IP address, record and timestamp.

Modern credential handling

Argon2id password hashing, magic-link invites, 15-minute access tokens with 30-day refresh.

Role-based access

Platform, tenant, course and group roles. Someone can author one course and be a learner in another.

AI governance controls

AI should speed up human work without taking control away. Every AI output is editable before it is published.

  • AI-generated labels
  • Human approval workflow
  • Source grounding
  • Source-only generation
  • Prompt & generation history
  • Unsafe prompt moderation
  • Tenant AI policies
  • AI provider selection
  • Retention policies
  • No training on client data
AI infrastructure

A dedicated AI orchestration layer.

AI orchestration

  • Multi-model support
  • AI provider selection
  • Vector search
  • Retrieval
  • Prompt templates
  • AI caching
  • Cost controls

Document intelligence

  • OCR
  • Document parsing
  • Image generation connectors
  • Moderation

Tenant AI policies

  • Retention policies
  • Unsafe prompt moderation
  • Prompt & generation history
  • AI usage monitoring
Operations

Run with confidence.

Observability

OpenTelemetry, Prometheus and Sentry, wired in from day one.

Platform

  • Encryption
  • Audit logs
  • Built to support GDPR obligations
  • Disaster recovery
  • High availability
  • Scalable microservices
  • Tenant-aware services

Access control

  • Platform roles
  • Tenant roles
  • Course roles
  • Group roles
  • Permission history
  • Password policies
  • JIT provisioning
  • Session management

Move from “learning delivered” to learning that delivers results.

See BlitzLIS with your own content. Our team will walk you through the AI course generator, multi-tenant setup and analytics.