Palpabl NMS gives IT teams real-time network visibility, config monitoring, and agentic alerts without the complexity of legacy tools. $39/mo.

Palpabl exists because self-hosting NetBox is a tax nobody wants to pay.

If you run a network, you've probably hit the point where the spreadsheet stops scaling and you need a real source of truth: device inventory, IPAM, DCIM, circuits, the whole graph. NetBox is the accepted answer. It's also a Postgres/Redis stack you have to stand up, secure, back up, and babysit through every upgrade and plugin-compat matrix.

For a small team, that upkeep is exactly *why* a source of truth starts to feel like overkill.

Palpabl is that same NetBox data model, hosted and managed. You get the source of truth without owning the infrastructure under it.

Being upfront about what it is and isn't, because this crowd tests every claim in session one:

What it does today:
- Network modeling, IPAM, DCIM, circuit + connection tracking — multi-site out of the box
- Config templates (Jinja2), webhooks, change workflows, compliance rules
- Drift detection (intended vs. actual), topology mapping, periodic network discovery
- A GraphQL API — one query pulls exactly the slice of the graph you want
- Agentic Triage: natural-language, multi-step **read** queries across your documented inventory
- Cisco Meraki API-key ingest, more integrations in progress

What it is NOT
- Not a monitoring tool. No uptime/latency/bandwidth graphs, no alerting. Keep your PRTG/LibreNMS/Zabbix — Palpabl is the source-of-truth layer those tools don't give you.
- Discovery is a periodic scan, not continuous polling.
- Triage reads your data; it doesn't push changes to live devices.

Free up to 20 devices (card required, the AI features bill at API rates, so we ask upfront rather than surprising you). Paid starts at $39/mo, full pricing on the site, no sales call to see it.

Built for the person who *is* the whole IT department, or the network engineer tired of maintaining the tool that maintains everything else.

Honest ask: poke holes. Tell us where the model doesn’t fit your environment, what integration you’d need before it’s useful, and whether “managed source of truth” is even a category you think in — we suspect most teams split managing and monitoring and never name the layer in between.

Setup was way faster than I expected, had my first config alerts within minutes. The agentic alerts actually catch stuff instead of just spamming me like my old monitoring tool did.

@elife1174609 thank you for your feedback! We do appreciate it!

How does the agentic alerting actually work in practice, like is it just doing threshold checks or can it correlate across devices to cut down on the noise during outages?

How does the alerting actually work in practice — is it just threshold based or can the agentic part correlate events across devices before pinging us?

How does the agentic alerting actually work in practice, like what triggers an agent to act versus just sending a regular notification?