Reliability & Security

Live platform health

SLO commitments

• • Availability target: 99.9%
• • API latency target (P95): < 400ms
• • Queue delay target (P95): < 10s
• • P1 first response target: < 15 minutes

Security controls

• • Role-based access control with tenant scoping
• • Platform-admin guardrails for global governance actions
• • 2FA enforcement for privileged roles
• • Request correlation IDs for support diagnostics
• • Rate limits on auth, SSO, billing, and key operations

Incident communication

We use a structured incident communication workflow: initial notice, regular progress updates, resolution summary, and preventive follow-ups.

For critical incidents, updates are issued on a scheduled cadence until resolution.

Where your data lives

Primary data is hosted on infrastructure in the European Union. Traffic is encrypted in transit over TLS, and the storage infrastructure uses full-disk encryption. Provider credentials, SMTP passwords and DKIM private keys carry an additional application-level encryption layer, and account passwords are stored as hashes, never in clear text.

Each customer data set is logically separated, and API keys never appear in logs. Retention periods are publicly documented and enforced automatically by scheduled jobs rather than left to anyone discretion: campaign messages are deleted after 60 days, individual messages after 30, and audit logs are kept for 365 days.

Governance and compliance

You are the controller of the data you upload and we are the processor. The relationship is governed by a publicly available data processing agreement that applies automatically, with no separate negotiation, covering security measures, subprocessors, international transfers and deletion of data on termination. In the event of a security breach affecting your data, we notify you without undue delay, with the information you need for your obligations under Art. 33 and 34 GDPR.

Continuity and backups

We run regular database backups with tested restore procedures and an explicit retention ceiling, so storage cannot grow unchecked. Processing queues are monitored continuously, and the team receives automatic alerts for failed jobs, delivery incidents or degradation in operational indicators.