Get in touch.

Nothing changes on your mail server. MxGuard sits in front as your MX record, scores incoming mail, and forwards clean messages to your existing server (Plesk, Microsoft 365, Google Workspace, anything that speaks SMTP). Quarantined and rejected mail never reaches your server.

Traditional anti-spam stacks rely on hand-tuned rule trees built over decades. They're slow to adapt to new spam and produce frequent false positives. MxGuard uses one trained ML model that learns from your team's own ham/spam labels, plus complementary threat-feed and reputation signals.

Quarantined messages are held in our queue until you release or delete them (auto-deleted after 30 days by default). Released messages flow through to your backend and aren't kept. Rejected messages are never stored. Body text used to improve the model is never shared or used outside MxGuard.

MxGuard runs on UK-based infrastructure operated by Transcom Internet Services (Brighton). All mail and metadata stays in the UK. GDPR-aligned by design.

Yes — we import your existing allow/block lists during onboarding and provide guided cutover assistance free of charge on any paid plan. eFa is end-of-life July 2026 — get in touch before then to avoid the rush.

No. The 30-day trial gives you full feature access with no card on file. We'll prompt for payment only if you decide to continue past the trial.

Sending servers retry for 3–5 days when an MX is unreachable — standard SMTP behaviour. Brief outages don't lose mail. A secondary MX (mx2) for high availability is being deployed during 2026.