Security Overview
Security is foundational to Tegendo.AI. As an enterprise AI platform handling sensitive organizational data and conversations, we implement defense-in-depth security across every layer of the stack — from network to application to data.Security architecture
Tegendo.AI is built on a modern, cloud-native architecture deployed on AWS with security controls at every layer:Infrastructure layer
- Hosting — Deployed on AWS with infrastructure spread across multiple availability zones for high availability
- Network isolation — All services run in private VPCs with no direct internet access. Only the API gateway and load balancer are publicly accessible
- WAF — AWS Web Application Firewall protects against common web attacks (SQL injection, XSS, DDoS)
- DDoS protection — AWS Shield provides automatic DDoS mitigation
Application layer
- Authentication — SSO via SAML 2.0 / OIDC, with optional MFA enforcement
- Authorization — Role-based access control (RBAC) with fine-grained permissions
- Session management — Configurable session timeouts, secure cookie handling, and automatic session invalidation
- API security — All API endpoints require authentication via API keys with scoped permissions
- Input validation — All user inputs are validated and sanitized before processing
- Rate limiting — Per-user and per-organization rate limits to prevent abuse
Data layer
- Encryption at rest — AES-256 encryption for all stored data (see Encryption)
- Encryption in transit — TLS 1.3 for all connections
- Tenant isolation — Row-level security ensures strict data separation between organizations (see Tenant Isolation)
- Data retention — Configurable retention policies with automated deletion
Data handling
What data we store
| Data type | Stored | Encrypted | Retention |
|---|---|---|---|
| User profiles | Yes | AES-256 | Account lifetime |
| Conversations | Yes | AES-256 | Configurable |
| Uploaded files | Yes | AES-256 | Configurable |
| Agent configurations | Yes | AES-256 | Until deleted |
| Audit logs | Yes | AES-256 | 90 days (default) |
| API keys | Yes (hashed) | bcrypt | Until revoked |
What data we do NOT store
- Model API keys — When using BYOK, your API keys are encrypted and never logged or exposed after initial setup
- Plaintext passwords — Passwords are hashed with bcrypt and never stored in plaintext
- Conversation content after retention — When the retention period expires, conversation data is permanently deleted
AI model data handling
When you send a message through Tegendo.AI, here is how your data flows:- Your message is sent to the Tegendo.AI API over TLS 1.3
- The message is processed by our backend to apply system prompts, skills, and context
- The composed prompt is sent to the selected AI provider (Anthropic, OpenAI, or Google) via their API
- The AI provider processes the request and streams the response back
- The response is streamed to your browser in real-time
- Both the request and response are stored in the conversation history (encrypted at rest)
Tegendo.AI uses enterprise API agreements with all AI providers. Under these agreements, providers do not use your data for training their models. Your conversations remain private and are not shared with or accessible to the AI providers after processing.
Access control
Authentication methods
- SSO (SAML 2.0 / OIDC) — Enterprise single sign-on with your identity provider
- Email and password — Standard authentication with bcrypt password hashing
- MFA — Optional multi-factor authentication via TOTP (authenticator app)
Authorization model
Tegendo.AI uses role-based access control (RBAC) with four built-in roles. Admins can create custom roles with fine-grained permissions. See the RBAC guide for details.API authentication
API access uses scoped API keys that can be restricted to specific operations. Keys are generated in the admin dashboard and can be revoked at any time. See the API overview.Audit logging
Every significant action in Tegendo.AI is recorded in the audit log, including:- User authentication events (login, logout, failed attempts)
- Conversation creation and deletion
- Agent creation, modification, and deletion
- Admin configuration changes
- API key creation and revocation
- Member invitations and role changes
Incident response
Tegendo.AI maintains a documented incident response plan:- Detection — Automated monitoring and alerting for anomalous activity
- Triage — Security team assesses severity and impact within 15 minutes
- Containment — Immediate actions to limit the scope of the incident
- Notification — Affected customers are notified within 72 hours per our SLA
- Remediation — Root cause analysis and permanent fixes
- Post-mortem — Documented lessons learned and preventive measures
Vulnerability reporting
If you discover a security vulnerability, please report it responsibly:- Email: security@tegendo.ai
- Response time: We acknowledge reports within 24 hours and provide an initial assessment within 72 hours