Sending Fleet State Machine

A cold email program is dozens of sending domains, hundreds of inboxes, and thousands of sends a day. Any one of those pieces can go bad — a domain lands on a blocklist, an inbox disconnects overnight, a campaign starts hitting a wall of bounces. The platform knows, but it doesn't tell you; the default state of deliverability information is buried. This is the place where it all surfaces: one dashboard the whole team can read, backed by automations that handle the routine responses on their own.

Each view answers one question. Is anything on fire? Are the domains healthy? Which inboxes are connected, sending, or quietly broken? Where are bounces concentrating, and how do replies break down by provider — because “we're fine on Google but dying on Microsoft” is exactly what averages hide. Per-domain health, per-provider reply breakdowns, and cohort comparisons find the sick patient in the ward instead of reporting the ward's average temperature.

The automations: every bounce gets classified, with hard failures and risky addresses going straight to the blacklist and a log of every action taken. Replies get sorted and routed so interested prospects surface fast. Daily safety sweeps sit behind every webhook, because webhooks miss things and inboxes fail quietly — the design assumption is that any single alert can fail, so every alert has a sweep behind it. And the fleet itself runs on its own clock: warmed ahead of demand, cleared by a human before campaigns draw from it, rotated as stock degrades. EmailGuard carries the placement and domain-health reads, with GlockApps, Warmy, and MailReach as the monitoring bench, and inbox stock drawn from InboxKit, HyperTide, ScaledMail, and Maildoso.