Where your data goes, and where it doesn't

Deployment

SmartNotifier runs on your infrastructure. Either as a Windows service on a server in your network, or as containers on your Docker host. There is no SmartNotifier cloud in the alarm path.

Data handling

• Snapshots and clips are stored in your storage, behind time-limited links, and removed on the schedule you set.
• Alarm and notification records live in your database — SQLite for a single site, PostgreSQL for larger deployments.
• Channel credentials are encrypted at rest and masked whenever they are read back through the configuration API.
• We never receive, proxy, or store your video.

Access

• Per-site accounts with hashed credentials.
• Token-based authentication on every configuration and audit endpoint.
• Plugin-to-backend traffic is authenticated per site.
• Configuration changes are recorded.

Network

The only outbound connections are to the messaging providers you enable, and a periodic licence check. The licence check caches for 24 hours and tolerates a week offline, so a network outage never stops alarms going out.

Resilience

• Queued and scheduled work, including pending escalation checks, survives a service restart. An escalation due in four minutes still fires if the box reboots in between.
• Every delivery is retried at the channel level and again at the job level.
• A monitor sweeps for alarms that have gone unhandled longer than expected and flags them.

Audit

Every alarm, every message, every acknowledgement, every escalation. Exportable.

For a monitoring or alarm-response operation, that is the evidence that an SLA was met — not a claim that it was. What the record contains.