MxGuard from Transcom

MxGuard is a cloud-based inbound email security gateway. Domains point their MX records at MxGuard, which scores every message and then rejects, quarantines, tags or forwards it to the customer's existing mail server. There is no hardware or software to install, and it works in front of any mail server.

Detection: each message is scored by a custom-trained LightGBM machine-learning classifier that evaluates subject, body, headers and links together, alongside DNSBL and URL reputation checks, SPF, DKIM, DMARC and ARC validation, ClamAV virus scanning and attachment risk analysis. Messages with an uncertain verdict are escalated to a large language model (Claude) for a second opinion. Scoring typically takes under 5 milliseconds per message.

Transparency: a live mail feed shows every message with its score and verdict, and each message has a detail view breaking down how the score was reached (ML model, URL reputation, sender reputation, authentication results and heuristics). Users can mark messages as ham or spam, and those labels are used to retrain the model.

Management: self-service domain onboarding with DNS ownership verification, 30-day quarantine with self-service release, daily digests, allow and block rules at four scopes (recipient, domain, customer and global), and a per-customer dashboard with volume trends, top recipients and top blocked senders.

Resilience: two gateway nodes in the UK and EU run active-active with replicated state, so mail continues to flow if either node is offline, with no DNS change required.

Outbound: an optional smarthost relay lets self-hosted mail servers send through MxGuard's infrastructure using SMTP authentication or an IP allowlist, with ClamAV scanning, rate limiting, optional DKIM signing and full logging. This addresses deliverability problems caused by poor IP reputation or blocked port 25.

Hosted in the UK and EU and GDPR-aligned. A 30-day free trial is available.