Deliverability autopilot

Most email APIs make deliverability your problem. NoticeAPI watches it for you: every bounce and complaint is tracked against rolling thresholds, and if your traffic starts hurting your reputation, sending pauses before Gmail and Outlook start filtering you, which is far easier to recover from.

For the product overview, read the deliverability autopilot guide.

What runs automatically

Auto-suppressionHard bounces, complaints, and provider blocks are suppressed account-wide instantly.
Health monitoringBounce and complaint rates over a rolling 24h window. Paid plans pause at stricter production thresholds; Free starts with a card gate and a slower ramp. You're emailed when a pause happens. Metrics explains the signal, cooldown, and exact recovery path.
Warm-up rampDaily limits grow day by day for new accounts. Free ramps more slowly and stays capped at 100/day; paid workspaces get more headroom as healthy traffic builds history.
Broadcast complianceMarketing sends always carry a working unsubscribe link and one-click List-Unsubscribe headers, so Gmail/Yahoo bulk-sender rules are satisfied by construction.

What it means for you

Watch the Metrics page: the charts show your rates against the RISK and PAUSE lines. If a send returns 403 project_paused, the current project crossed a threshold. Sandbox remains available. First-time bounce incidents move automatically from cooldown into small production probes, then regain capacity as real outcomes stay healthy. Spam complaints, repeated incidents, workspace emergency pauses, and ambiguous legacy pauses still require operator review.

Self-service recovery

During a bounce cooldown, Metrics shows the active signal and next evaluation time. Once the risk window clears, NoticeAPI opens a small server-controlled production budget. Healthy probe outcomes expand that budget in stages until normal capacity returns. Customers cannot force a restart; the observed delivery evidence controls every transition and each transition is written to the audit trail.

Why we're strict

Sending infrastructure is shared. One customer blasting a scraped list degrades delivery for everyone on the platform, so we'd rather pause a bad list in minutes than let it burn a domain for weeks. The thresholds are tuned to never trigger on legitimate transactional traffic.