Built for board-level trust.
CIOs entrust Kogira with their organisation's most sensitive strategic documents. Here is exactly how that trust is protected — control by control.
Assurance Statement
Kogira processes your strategy documents and organisational evidence exclusively through server-side infrastructure. Your content is never exposed to AI providers via browser-side calls, never retained by our AI provider beyond the response window, and never stored in plaintext at any layer. All access is authenticated, rate-limited, and schema-validated before a single token is processed. This posture has been designed to satisfy the security requirements of regulated industries including financial services, healthcare, and government.
Data Sovereignty
Where your data goes — and where it doesn't.
Your data never passes through our servers to AI providers
Kogira routes all AI processing through isolated server-side functions. The AI provider never receives your IP address, session token, or identity — only the content you explicitly submit.
API credentials are never exposed to the browser
The connection to the AI provider lives exclusively in a sandboxed server environment. No credential ever appears in a browser, network response, or client-side bundle.
All data in transit is encrypted with TLS 1.3
Every request — from your browser to Kogira, and from Kogira to any external provider — travels over TLS 1.3. There are no unencrypted hops at any stage of the pipeline.
Zero-day data retention policy with our AI provider
Kogira operates under a zero-day data retention agreement with its AI provider. Your prompts and strategy content are not retained, logged, or used for model training beyond the response window.
Access & Tenant Isolation
Who can access what — and the guarantees that enforce it.
Row-Level Security on every table
Every database table enforces Row-Level Security policies. Your data is isolated at the database layer — not just application logic. Even a misdirected query cannot cross tenant boundaries.
Rate limiting and abuse controls enforced per organisation
Each organisation is subject to per-minute, per-hour, and burst rate limits on authentication and API endpoints. This prevents cost exposure from abuse and protects platform availability.
Multi-factor authentication and role-based access
MFA via TOTP and SMS, with three-tier role-based access control — platform admin, company admin, and standard user. All access changes are logged to an immutable security audit trail.
Security enforced at build time, not just runtime
Custom static analysis rules run in CI on every pull request, enforcing tenant isolation patterns and preventing API keys or direct AI calls from appearing in client-side code.
Resilience & Governance
What happens when things go wrong.
Credits are deducted atomically — no bypass possible
Platform credit consumption uses atomic database operations. There is no race condition that could allow usage without corresponding deduction.
The platform handles provider degradation gracefully
Circuit breakers and exponential backoff with jitter protect against cascading failures during AI provider degraded conditions. Your requests queue and retry safely.
Data at rest is encrypted
All data stored by Kogira — documents, assessments, strategy outputs — is encrypted at rest using AES-256. There is no plaintext persistence at any layer.
Security events are logged without storing personal data
Platform security events are logged using hashed identifiers, not raw IP addresses or personally identifiable information. Audit trails are complete; personal data exposure is not.
Monitoring & Compliance
How we detect, respond, and prove it.
Real-time error monitoring with Sentry
Every unhandled exception is captured automatically with full stack trace, user context (no PII), and session replay. The platform operator is alerted by email on every new issue.
24/7 uptime monitoring with phone alerts
An independent external service checks platform availability every few minutes. If the health-check endpoint fails, the platform operator receives an immediate phone call.
Passkey authentication supported
Kogira supports passkeys (WebAuthn) including biometric methods such as Face ID and fingerprint — the most phishing-resistant authentication method available today.
SOC 2 Type II — controls operating, observation period in progress
Kogira's security controls are designed and operating to SOC 2 Trust Services Criteria covering Security, Availability, and Confidentiality. Nine formal policy documents govern access control, change management, monitoring, incident response, risk assessment, and vendor risk. Our infrastructure partners — Supabase and Vercel — hold current SOC 2 Type II reports. Stripe is PCI DSS Level 1 certified.
Ready to see how Kogira protects your strategy?
No credit card required. Upload your first document free.
Questions about security? Contact us directly