Everyone in one workspace
Coworkers on your company domain join automatically — no per-seat setup, no per-seat pricing. What one teammate connects, the whole team can ask about.
Connect →Correlate & RCA →Autopilot
Connect your whole network once — then find the root cause.
Sauble hosts and validates the connectors — Catalyst Center, SD-WAN, Prometheus, Elastic and more — so your systems are AI-ready with zero servers to run. Ask in plain English and Sauble correlates every connected source to a likely root cause, with the evidence behind it. When you're ready, it investigates on its own.
No card required · Free starter credits · Coworkers on your domain share one workspace
✻ Sauble connected · 7 sources across Portland
>Why do clients at the Portland office keep dropping off the network?
⏺sauble - catalyst_center(site: "Portland")
⎿3 APs flapping — ap-pdx-03/07/11, repeated deauth events
⏺sauble - prometheus(query: "uplink errors core-sw1")
⎿ifInErrors rising on Gi1/0/1 over the last 2h
⏺That failing uplink is the likely root cause. Those APs uplink to core-sw1, which shows rising errors on its core link. Recommended: check the transceiver on core-sw1 Gi1/0/1.
Connectors · Connect
Connect the tools you already run.
Point Sauble at your own monitoring, network, and observability systems — we host and validate every connector, so there are no MCP servers for you to stand up. Read-only by design.
New connectors added every week Browse the full catalog →
Collaboration · one shared brain
Not a personal assistant — a brain your whole team shares and teaches.
Connect a source once and everyone on your domain can ask about it. The context your team builds — network descriptions, past investigations, runbooks — lives in the shared workspace and makes every future answer sharper. Knowledge stops walking out the door with whoever was on call.
Knowledge that compounds
Network context, maintenance windows, baselines and past investigations accumulate in the workspace — so Sauble answers with your team's hard-won knowledge instead of starting from scratch each time.
Shared sources, context & credits
Sources, network context, memory and credits are pooled across the team. Set it up once; everyone works from the same source of truth — and the same running total.
How it works
Connect → Correlate & RCA → Autopilot
The same three beats are the pitch and your path through the product. What you build in the Playground is the same brain that goes autonomous in production.
Connect
Pick your systems — Catalyst Center, SD-WAN, your own MCP server. Hand Sauble a read-only credential and we run the connector. You deploy nothing.
Correlate & RCA
Ask in plain English from Claude Code, Cursor, or Sauble. It correlates controller state, telemetry, topology and logs across every connected system, then explains the likely root cause — with the evidence and confidence behind it, not just an answer.
Autopilot
The same brain your team queries by hand runs the investigation itself the moment an alert fires — root cause, impact, and evidence ready before an engineer looks.
Why Sauble
Four things that are genuinely hard — done for you.
Connectors lead, but each of these is real engineering effort you don't have to take on.
1
Connected in minutes
Your network systems, AI-ready — no servers to run. Pick a connector, give it a read-only credential, and Sauble hosts and validates the bridge. Already run your own MCP servers? Point us at them.
2
Root cause, explained
Sauble doesn't just retrieve — it reasons. Symptoms scattered across controller state, metrics, topology and logs become one incident narrative with a ranked likely cause, the supporting evidence, and blast radius. That correlation is the moat.
3
One shared brain
Coworkers on your domain join automatically. Sources, context, knowledge and credits are shared — what one teammate connects or documents, everyone can ask about. No per-seat setup or pricing.
4
Governed by design
Sauble sits between your systems and any model: credentials are brokered and never reach a model, only what an investigation needs leaves your systems, and every action lands on the audit trail.
See it work
From zero to answers in under a minute
Sign up, save your key, connect your sources, and ask. Tap any step to jump straight to it.
sauble.ai/playground
Pricing
Start free. Scale by credits, not seats.
Everyone on your domain shares one pool of credits — no per-seat pricing. You pay for the investigations you run, not the number of people running them.
Playground
Freeno card required
500 tokens of starter credit
Try Sauble against a sample network or your own sources — non-production use.
- Shared workspace on your domain
- Connect your own sources or a sample network
- Ask from Claude Code, Cursor, or Sauble's UI
- Community support
Team
1–10seats · one shared pool
X credits / month, pooled across the team
Production use for a whole team working from one shared brain.
- Everything in Playground
- Production use across your team
- Pooled credits — no per-seat pricing
- Priority connector requests
- Email support
Enterprise
Customtalk to us
Custom credits & seats
For large, multi-team operations with security and deployment requirements.
- Everything in Team
- Autonomous alerts at scale (Autopilot)
- Private deployment & data controls
- SSO, audit trail, custom connectors
- Priority support & SLA
Demo
See the investigation flow in action.
One brain for your network. Find the root cause today. Let it run tomorrow.
Connect a sample network in one click, or point Sauble at your own gear in five minutes — free, no card required.