DeviceWatch
SNMP monitoring where AI learns what's normal per interface
Start new thread

DeviceWatch - SNMP monitoring where AI learns what's normal per interface

by•
AI-powered SNMP network monitoring software for macOS, Windows and Linux. Monitors switches, routers, firewalls and servers — on-device AI learns your traffic baseline and alarm habits. Up to 250 devices. No cloud, no agents, no per-node licensing.

Add a comment

Replies

Best
Maker
📌
Hey Product Hunt — I'm the developer. DeviceWatch polls your switches, routers, firewalls and servers over plain SNMP v2c/v3. No agents, no collector to stand up, no cloud account — it runs on one machine and talks straight to your gear. The part I'm actually proud of: on-device AI builds a per-interface baseline (time of day + day of week) instead of one static threshold for everything, and it learns which alarms you keep dismissing so it stops emailing you about those specifically — it never silences an actual device outage, though. Rather than describe it, try the demo, no signup: meshwatch.app/demo/devicewatch/ Go to Alarms, switch "All alarms" to "Grouped by cause", and watch 14 alarms collapse into 2 actual incidents. That's basically the whole pitch in one click. macOS/Windows/Linux, up to 250 devices per licence ($999/yr), 30-day free trial with everything unlocked, no card required. Network maps are built from LLDP/CDP. Honest limits: the Windows build isn't code-signed yet, so you'll see a SmartScreen warning on first run. Hardware health profiles are field-tested for most major vendors, but Juniper and Cisco WLC are implemented and explicitly flagged in the app as not yet verified against real hardware. Happy to answer anything — what would make this useful in your environment?