For SolarWinds Observability Self-Hosted

AI agents for the SolarWinds you already run

Your team asks a question and gets an answer about the whole environment. And it doesn't stop there. The agents investigate and can do the work too, with your approval.

Nothing gets copied out of SolarWinds, and nothing gets replaced.

Your tools
SolarWinds Self-Hosted
NPM NCM NTA IPAM
Used by NetOps + AI SRE
Other ITOps tools
Ticketing App monitoring Logs Cloud
Used by AI SRE
Fabrix.ai agents
One Ubuntu VM in your environment
NetOps agents AI SRE agents
Investigate Explain Act, on approval
Your AI account Cloud provider Fabrix-managed On-premises

Two sets of agents. Start with the first.

If SolarWinds is all you need covered, NetOps agents work with nothing else connected. If your incidents also run through other tools, AI SRE agents follow them there.

When you run SolarWinds on its own

NetOps Agents

The data is already in SolarWinds and your team knows the platform. The time goes on the work in between: understanding one alert, checking whether a config change caused it, working out what it affects. These agents take that on using only what SolarWinds already sees.

SolarWinds Insights Agent
Ask anything about the environment. No query syntax.
Network Alert Impact Agent
What an alert actually affects, using topology and inventory.
SolarWinds RCA
Root cause, investigated and written up with evidence.
Asset Config Management Agent
What changed, and what a change would break.
SACM Assessment Agent
Where your monitoring has gaps, and how to close them.
CVE Agent
Which CVEs touch your live inventory, ports and configs.

More are built for whichever modules you run. Admins also get AI Observability for agent health and spend.

View NetOps Agents Solution Brief

When SolarWinds sits alongside other tools

AI SRE Agents

An incident rarely sits in one tool. The alert is in SolarWinds, the symptom is in app monitoring, the change is in ticketing and the evidence is in your logs. Your engineers stitch that together one console at a time. These agents run one investigation across all of it.

Cross-Domain RCA Agent
One investigation across SolarWinds, logs, ticketing and app monitoring.
Service Impact Agent
Which services and users an infrastructure event actually affects.
Incident Context Agent
Enriches a ticket with topology, changes and related alerts.
Change Correlation Agent
Whether a recent change explains what is happening now.
Access Diagnostics Agent
One user's connectivity problem, end to end.
Health Sweep Agent
A scheduled posture check across the whole stack.

Works with ticketing and ITSM, app monitoring, logs and data lakes, infrastructure and cloud, and security tools. More agents are built around the tools you actually run.

View AI SRE Agents Solution Brief

Whichever you start with

Answers in minutes
The routine first hour of an incident is done before anyone picks it up.
Launch it, don't prompt it
Run it on demand, on a schedule, or let an alert trigger it.
No data migration
Agents read each tool where it sits, through read-only APIs.
You approve changes
Anything that changes state waits for a person. Every run is audited.
Your data, your model
Local model or cloud. Your policy decides.
One Ubuntu VM
No device agents to roll out and no changes in Orion.

Questions we usually get

Doesn't SolarWinds already have AI?

Its AI features are part of the cloud-hosted SaaS product, so Self-Hosted customers don't get them. Our agents are built only for Self-Hosted. They sit alongside SolarWinds rather than competing with it.

Do we have to move our data?

No. The agents read SolarWinds where it sits. There's no copy and no migration into a new platform.

Can we trust it to act?

It's read-only by default. Any action that changes state needs a person to approve it, and there's a full audit trail on every run. If you pick an on-prem model, no data leaves your environment.

We already use Claude. Isn't that enough?

Keep using it for coding. This is for running the network. A copilot runs when someone types, and it has no live connection to NPM, NCM, NetFlow or logs, so every investigation turns into copy and paste. These agents run on a schedule or on an alert, unattended.

How is it priced?

It's an annual subscription: a platform tier plus AI usage credits, which still apply if you bring your own model. We scope it with you on the first call.

Want to see it in action?

Start with NetOps on the SolarWinds you have today. Add AI SRE when you want one investigation to reach the rest of your tools.

Request a Demo

Fabrix.ai agents are a Fabrix.ai product, not a SolarWinds product.