The Agent Team

Generic AI tools reason about networks using textbook patterns. NetBrain's AI agents reason about yours: your circuit naming, your escalation paths, your network's own history.

Meet them through AI Copilot, one chat interface for the whole NetBrain Agent Team. Describe what you need in plain language, and the right agent takes it from there. Every agent reads before it writes. Every action that matters waits for your approval.

System Agents

Built-in, provided by NetBrain. Customers can edit Resources, Skills, and MCP Servers, but not the System Prompt or locked built-in tools.

Deep Diagnosis Agent — Assists with root-cause analysis. Triggered by ticket, alert, or webhook; surfaces findings and hands off to Change or Assessment.
Assessment Agent — Proactively checks the network by turning a plain-language requirement into an Intent, ADT, and dashboard.
Change Agent —Generates and assembles Change Runbooks without executing the network change itself.
Companion Agent — A default everyday agent that maps operations, performs runbook analysis, DB queries, CLI execution, Intent creation. It routes to Deep Agents when needed.

AI Copilot

AI Copilot gives your team one interface to ask your network anything in natural language: a diagnosis, an assessment, a change, or just a quick question.

Ask a question and Copilot routes it to the right agent: Diagnosis Agent for root cause, Assessment Agent for checks across the network, Change Agent for governed runbooks, or Companion Agent for everyday questions. Every session starts with Companion ready to go.
If your prompt shifts to a different agent’s domain, Copilot switches automatically and notifies you inline, no extra step needed. If a workflow suggests handing off to another agent instead, like Diagnosis to Change, Copilot surfaces an Action Card for your approval first.
Within the Assessment Agent, AI Parser builds a working parser from a plain-language request: tell it what you want, and it identifies the device type, picks the right command, and generates the parser. It covers CLI, Config (including Juniper), and API parsers, and can be previewed, edited, and saved by junior engineers and non-technical users alike.
AI Document uses AI to generate network documentation automatically from live operational data. NetBrain’s AI agents summarize network activities, translate operational findings into narrative documentation, and export Observability Dashboard content directly to Word or PDF. The agent reads the context-aware digital twin and the activity log, then writes the documentation the team would have written if they’d had the time.
- Auto-generate documentation for network diagrams, device configurations, and operational processes.
- Export in editable formats including Microsoft Word and Microsoft Visio.
- Include design and inventory data, diagnosis details, full configuration files, and routing tables.
- Translate Observability Dashboard snapshots into a narrative report ready for leadership.
- Shareable, audit-ready, and current.

Deep Diagnosis Agent

The Deep Diagnosis Agent finds root causes for complex, novel incidents across multi-vendor networks. Triggered by a ServiceNow ticket, an alert, or a webhook, it navigates live network topology, runs diagnostics, and hands you a structured root cause and recommendation.

See its reasoning transparently as it happens. When it needs more context, it asks for it instead of guessing.
Handing off the result is built in: Remediate sends it to the Change Agent, and Assess Similar Devices sends it to the Assessment Agent. No re-explaining the problem to a second agent.
No unsupervised fixes. Every recommendation waits for engineer review or a pre-approved remediation.

Assessment Agent

The Assessment Agent turns a plain-language question into a full assessment. You describe what you want to check, and it builds and runs an Intent, an ADT, and a dashboard behind the scenes.

Scope the check to a device group, a site, a set of map devices, or your whole domain.
The agent finishes the assessment before it ever offers a dashboard, even if you asked for one up front. You get the button; you decide whether to click it.

Change Agent

The Change Agent turns a change goal into a governed runbook: pre-check, execute, post-check, compare, document. You never touch a device directly, and neither does the agent.

Every runbook sits in review until you approve it. Devices outside the change scope are excluded automatically.
Start by typing your goal directly, or hand off straight from a Deep Diagnosis or Assessment result.
AI-generated changes carry an “AIBot” tag, so your audit trail always shows who, or what, proposed the change.

Companion Agent

The Companion Agent is your everyday agent in AI Copilot, handling routine requests and routing anything else to the right specialist. Ask it a question, and it can create a Network Intent straight from the conversation, run a query and render the answer as an interactive chart right there in the chat, or execute CLI commands and Intent or Runbook nodes directly.

Ask it to draw topology on a map, extend neighbors with live data, or visualize compliance checks as overlays. Mid-conversation, it can hand off to Deep Diagnosis without losing context.
Outside its own scope, it hands off to Diagnosis, Assessment, or Change instead of forcing an answer.

AI Doctor

NetBrain’s Discovery and Benchmark processes collect every data-accuracy issue they find into one worklist. AI Doctor works that list: it resolves what’s deterministic on its own, applies your best stored credentials where a device needs authentication, and diagnoses root cause with a support ticket for anything a fix won’t make stick.

Anything that reshapes topology or deletes records waits for your approval by default.
Unlocking a locked device always needs your explicit consent, and the system, not AI, restores the original lock once the run completes.
Issues rooted in a driver-code gap, where any fix would be reverted by the next Benchmark, get diagnosed to root cause and packaged into a NetBrain support ticket rather than resolved directly.

AI Knowledge and Memory Studio

AI Knowledge and Memory Studio is designed to turn a pile of old tickets into automation you can reuse. Upload your ticket data and a time range of Deep Diagnosis results, and the agent is planned to cluster them into ticket types, then write a root cause summary and a diagnosis procedure for each one.

Planned: review each type before publishing, edit a root cause, merge duplicate types, or delete one that doesn’t hold up.

Every agent runs on the same foundation: Live Network Context for network state, Network Intent & Spec for what “correct” means, and Agent Services & Skills for the knowledge that’s specific to you. Design Studio is where your team sets the rules each one operates by.