Artificial Intelligence

AI Bot

AI-Driven Network Insights on Demand

Troubleshoot in natural language, ask diagnostic questions and receive context-rich responses, dynamic maps, and action plans for streamlined issue resolution.

Everyone’s Chatbot for Simple Automation Requests

Your translator and interface between engineers and their networks. Use natural language input to easily access network data like IP addresses, DNS, neighbors, and device logs without using CLI or a complex network management tool. AI will execute the automation for you and summarize the output in natural language in any format.

Shift-Left Troubleshooting with GenAI

Unlock the power of instant network configuration validation and auto-remediation. Our AI-driven solutions ensure that your network operates seamlessly, allowing for immediate detection and correction of issues as they arise.

Instant Network Configuration Validation and Auto-Remediation

Experience real-time monitoring that validates your network configurations instantly. With our auto-remediation capabilities, you can resolve potential issues before they impact your operations, keeping your network robust and reliable.

Automate Complex Troubleshooting with AI-Powered Action Plans

Say goodbye to cumbersome troubleshooting processes. Our AI-powered action plans simplify complex problem-solving, enabling your IT teams to focus on strategic initiatives rather than getting bogged down in technical difficulties. Enjoy more efficient operations and enhanced network performance with our cutting-edge technology.

Diagnosis, Change, and Assessment

The AI Bot assists engineers by providing actionable insights during core workflows. It answers network-specific “how-to” questions with tailored action plans and leverages the Digital Twin data model for optimal results.

Orchestrated Automation

Next, the AI Bot rapidly executes automation and multi-vendor CLI in conjunction with the no-code automation in the background.

Summarized Output

It presents summarized output in natural language, providing users with easily scannable tabular data and easy-to-interpret summary dashboards of automation results.

AI-Assisted Ticket Analysis

AI automatically processes, understands, and analyzes historical ticket data, teaching the automation and human how to troubleshoot similar incidents proactively in the future.

Streamline network operations, accelerate problem-solving, and build a more resilient network by turning your historical ticket data into actionable intelligence.

Intelligent Ticket Processing

• Advanced language models automatically read, interpret, and categorize incident tickets with unparalleled speed and depth.
• Understands the context and nuances of each reported issue.

Automated Knowledge Generation

• Automatically extracts useful information from historical tickets, turning raw data into a valuable, living knowledge repository.
• Generates actionable knowledge documents which summarize troubleshooting steps, making automation accessible and repeatable.

Accelerated Incident Resolution

• Significantly improves incident management efficiency by providing instant insights, directly contributing to reduced Mean Time to Resolution (MTTR).
• Standardizes troubleshooting processes, ensuring consistent and effective problem-solving across your team.

Proactive Problem Prevention

• Enhances triggered automation by learning from past incidents and identified patterns.
• Empowers your teams to shift from reactive firefighting to proactive outage prevention, stopping issues before they impact services.

Scalable Insights & Analysis

• Process massive volumes of historical data with powerful batch analysis capabilities.
• Creates “insight prompts” to accelerate diagnosis of similar incidents in the future.

How it Works

• Upload: Upload past ticket data in a CSV or Excel file.
• Analyze: AI automatically categorizes issues and finds solutions.
• Learn: Generates troubleshooting guides and prompts to help your team solve similar problems faster in the future.

Gain insights needed to troubleshoot incidents at an unprecedented scale.

AI Insight

Your Powerful Digital Operations Assistant

Analyze and summarize any mapped area of your network using AI to automate manual tasks and checks. Gain faster root-cause diagnostics and exponentially enhance network intelligence by leveraging results from tens of thousands of automations simultaneously. Maintain control as the senior engineer while the AI, guided by your expertise, conducts reasoning to answer critical questions about your network.

NetBrain’s AI-driven insights help you tackle a wide range of network challenges with ease. For example:

• Diagnose why R1 interface e0/1 is down or determine why host10.10.234.1 is unreachable.
• Investigate why R10 BGP neighbor 20.20.20.1 is down or identify the cause of Application XYZ runningslow.
• Verify multicast configurations by checking multicasting 239.0.0.1 or assess the health of site BCD.

• Ensure compliance by asking, “Is R10 NIST-compliant?” or review recent changes with, “Does R10 have any config changes last week?”
• Validate configurations by checking configuration rules for R10 or assess security risks with, “Does any device in the map have a CVE vulnerability?”

Puts Automation at Your Fingertips

Find and access all automation related to a mapped area of your network in a single place, including golden intent, golden config, ADT intent, intent from the observability dashboard, published intent, CLI, and config.

AI Runbook Companion

Stop wasting time trying to interpret data. Start diagnosing.

A workflow embedded AI companion that helps plan and recommends actions keeping engineers in control but out of the time-consuming investigative work. Empowers every engineer with the expertise and practical experience of a senior specialist, turning complex operational tasks into clear, actionable insights. Allows engineers to focus on core issues earlier, significantly reducing MTTR.

The Reader

Saves time reading raw outputs, correlating results across steps, and deciding what to do next. Runbook Companion compresses that cycle by converting runbook evidence into clear summaries, answers, and recommended next actions so engineers can move faster and more consistently from results to decision to action.

The Automation Recommender

Recommends suitable automations for cases without resolving the network problem, helping users apply the company’s best practices.

The Summarizer

Generates concise natural-language summaries with initial conclusions. AI proactively recommends the most appropriate automation tools—including CLI commands and NetBrain automation resources—and translates senior-level expertise into clear, actionable guidance.

AI Runbook Companion can:

• Analyze runbook outputs and return readable summaries and structured tables
• Answer questions about results, including anomalies, scope of impact, and suggested next checks
• Recommend suitable automations to continue troubleshooting or validate a change workflow

For the NoC Engineer in a Hurry
No more manually reading through CLI outputs, comparing tables, or deciding what to check next. Ask questions in plain English and get clear answers—right inside the runbook.

Example prompts you can use today:

• “Summarize the alerts in this runbook.”
• “Show batch ping results as a table.”
• “Why is voice quality poor between these endpoints?”
• “What changed on these devices in the last 2 hours?”

For the L2 and L3 Engineers
When a runbook doesn’t fully resolve the issue, AI recommends specific, best-practice automations to apply next including CLI checks, Golden Configs, Intents, or full Runbook Templates, keeping every engineer aligned with expert workflows.

For Leaders Tracking MTTR Efficiency
Get junior engineers to act like seniors. Keep troubleshooting consistent. Move from results to action in minutes, not hours.

What Makes It Different

• Context-Aware
  Reads runbook data (nodes, results, device outputs) and understands what you’re working on.
• Action-Oriented
  It doesn’t just summarize; it recommends exact next automations to run.
• Integrated
  Works inside NetBrain runbooks. Add AI suggestions as nodes with one click.
• Learnable
  Pin frequent prompts as “Goals,” share chats across teams, reuse expert questioning.

Real-World Scenarios

AI Deep Diagnosis for Complex Network Problem Diagnosis

Human-on-the-Loop Agentic NetOps Keeps Engineers in Control

The Investigation Engine: Where Agentic AI Meets Automation

Deep Diagnosis is an agentic AI capability that iteratively investigates across the live digital twin and intent-based automation to identify root causes fast.

• Analyzes issues with clear, step-by-step reasoning and visually maps the root cause for fast resolution.
• Enables junior staff to work like experts by turning complex troubleshooting into guided, automated steps.
• Keeps engineers in control with human-in-the-loop oversight and approval.
• Reduces resolution time dramatically by eliminating manual triage, distributed tools, and guesswork.
• Provides actionable conclusions backed by evidence, so teams can review and act with confidence.
• Turns proven AI investigations into reusable runbooks to operationalize fixes.
• Identifies root cause in over 90% of real-world network issues, accelerating the entire investigation cycle.

A Powerful Digital Operations Engineer

Deep Diagnosis isn’t just another analytics tool; it’s your autonomous, reasoning-driven partner. It ingests network context from the live digital twin, executes iterative checks, and delivers evidence-backed conclusions. This allows you to maintain control while the AI, guided by your expertise, handles the heavy lifting of root cause analysis.

Design Principles: Built for Trust and Control

NetBrain AI is built on three explicit principles that ensure it is safe, defensible, and aligned with how real network operations teams work:

• Operational Truth Over Inference: AI conclusions must be grounded in the live digital twin, real device data, and executed diagnostics. The AI is not permitted to invent explanations or operate on assumptions alone.
• Intelligence Without Unbounded Authority: The AI can decide what to investigate, execute diagnostics, and interpret results—but it cannot execute remediation independently or bypass change control workflows.
• Human Accountability Is Preserved: AI accelerates engineers; it does not replace them. Every AI-assisted action is attributable to a user, bound to a workflow, and governed by role-based permissions.

Autonomous, Reasoning-Driven Workflow

1. Trigger: A ticket, alert, or manual request kicks off an AI-driven investigation. The system understands the network context from the live digital twin.
2. Investigate: The AI acts as an autonomous agent, running iterative checks and gathering evidence across devices and paths to diagnose the issue.
3. Conclude: Get a clear root-cause conclusion, supported by step-by-step reasoning and visual evidence on a map and in Data Views.
4. Operationalize: Convert the AI’s investigation into a reusable, executable runbook—capturing and scaling the diagnostic workflow.

Inside the Agent System: How Deep Diagnosis Reasons

Deep Diagnosis is not a static script. It is a multi-agent system that dynamically determines how to solve your network problem.

• Triage Agent: Classifies your intent and determines whether the request requires diagnostic handling or map-related processing.
• Deep Diagnosis Agent: Performs autonomous reasoning and determines the evidence required to validate the troubleshooting intent.
• Retrieve Agent: Collects automation data through controlled, read-only operations from the relevant devices.
• Summary Agent: Produces a structured, human-readable summary without adding new reasoning, ensuring traceability.

Each agent executes a clearly defined stage of the workflow, ensuring modular, predictable, and traceable diagnostic behavior.

Diagnose Complex Problems Faster, With Less Effort

• Cuts Troubleshooting Time: AI reduces manual investigation from hours to minutes, delivering evidence-backed conclusions.
• Empowers Every Engineer: Junior staff follow expert-level reasoning, while seniors maintain oversight and control.
• Builds Institutional Knowledge: Turn proven AI investigations into reusable runbooks—so fixes are consistent and repeatable.
• Works Proactively and Reactively: Triggers automatically from tickets or runs on demand for immediate issues.
• Proven in Real Networks: Identifies root cause in over 90% of network issues, accelerating the entire resolution cycle.

Use Cases

Leverage Deep Diagnosis to tackle a wide range of network challenges with ease. Simply ask your question in plain English.

The Deep Diagnosis Experience: From Query to Resolution

1. Initiate the Investigation

Start a session from the desktop left bar. You can run queries with or without a specific map context. If you select Use map data, the AI scopes its analysis to the devices on that map, ensuring context-aware results. Each diagnosis session can contain multiple, independently running queries.

2. Review the AI-Generated Results

Once the diagnosis is complete, the AI presents a comprehensive findings report.

• Context Map: An auto-generated map visually pinpoints the issue. Relevant data is summarized directly on the map via Data Views.
• Conclusion: A clear statement identifying the root cause and its business impact.
• Key Findings: The critical evidence that supports the conclusion.
• Diagnostic Checkpoints & Results:A transparent, step-by-step breakdown of the AI’s investigation path.
• Recommended Actions: Clear next steps to fix the issue and verify resolution.
• Reference Links: All automations used in the diagnosis are listed, color-coded by status (Red = Alert, Green = Success, Blue = Informational).

3. Validate the AI’s Reasoning

Trust, but verify. Click the Reasoned for [time] link to see the AI’s complete thought process. This chain-of-thought reasoning provides full transparency, allowing you to understand how the conclusion was reached.

4. Operationalize into a Runbook

Move from automated diagnosis to guided remediation. The Add to Runbook button seamlessly transfers all evidence, findings, and automations into a new Runbook. The Runbook auto-groups CLI commands, intents, and configlets, creating a shareable, repeatable workflow for this fix.

5. Manage and Share Insights

All diagnoses are saved and organized under My Diagnoses and Shared Diagnoses. You can add custom descriptions, rename sessions, and share results with your team. The Export function generates a document of the entire diagnosis, while Share to Incident allows you to attach findings directly to a ticket with AI-summarized notes.

Centralize Your Diagnostic Power: The Deep Diagnosis Manager

The Deep Diagnosis Manager is your control center for enriching the AI’s knowledge. By connecting various data sources, you ensure the AI has the specific context of your network to perform accurate reasoning. This centralized governance model ensures consistent behavior and auditable configuration across all AI features.

Key Data Sources You Can Configure:

• Automations: Embed Golden Intents, ADT Intents, Published Intents, and Intent Templates.
• Runbooks & Configs: Add Runbook Templates and Golden Configs.
• Contextual Data: Link Maps, Device Groups, and Application Paths.
• Institutional Knowledge: Upload Knowledge Documents for proprietary scenarios (e.g., “The Omaha Test”) that the AI wouldn’t otherwise know.
• Parsed Data: Leverage the CLI and Config Dictionaries for efficient, cost-effective analysis of device outputs.

The database updates automatically every 4 hours, or you can trigger a manual refresh to make new automations available immediately.

Proactive Diagnosis in Your Workflow: Triggered & ITSM Integration

Deep Diagnosis is fully integrated with the Triggered Automation Framework (TAF). Now, when an incident is created from ServiceNow or another ticket system, an AI-driven diagnosis is triggered automatically.

Results appear directly in NetBrain Incidents, the Incident Portal, and the ServiceNow ticket, providing an instant AI-generated summary and a View Full Diagnosis link to explore the complete analysis in NetBrain. This shifts your team from reactive monitoring to proactive, AI-accelerated resolution—all while preserving audit trails and avoiding unsupervised remediation.

What Makes It Different

• Agentic, Not Just Assistive: AI doesn’t just summarize—it investigates, reasons, and concludes autonomously using a multi-agent system.
• Network-Aware Context: Uses the live digital twin to understand topology and dependencies.
• Transparent Reasoning: Shows its work—reasoning chain and evidence—so engineers can validate and trust the results.
• Constrained, Not Autonomous: Operates within safe, governed boundaries. It can investigate and recommend, but it cannot remediate without human approval.
• Action-Oriented Output: Delivers clear conclusions and converts them directly into executable automations.

AI Document

Adds Finesse to NetOps Reporting

Create a report with AI refinement using natural language input for tailored output.

Parser Assistant by AI

AI Parser Assistant helps teams create parser rules using guided prompts and AI assistance, reducing reliance on specialized parsing expertise and accelerating time to value for automation workflows.

AI-Powered Parser Creation

Generate parser rules in seconds without mastering complex grammar or regex. Simple AI prompts transform raw CLI output, including configurations, tables, status and outputs, into actionable, structured variables, tables, and paragraphs ready for automation, dashboards, and downstream systems. Instantly generate complete parser rules without writing or debugging parsing logic, keeping you in control to review, edit, and fine-tune the rules.

Parse CLI Output without Confusing Grammar

• Convert raw CLI output into clean, structured data instantly without needing visual parsing grammar.
• Use simple prompts to generate accurate parser rules, eliminating dependency on specialized expertise.
• Accelerate automation onboarding and standardize data extraction across teams.
• Turn device outputs into automation ready inputs for runbooks and assessments in less time.
• Validate and refine parsing logic quickly to support faster iteration.

Intelligent Rule Conversion

The system automatically translates AI-generated regex into standard NetBrain visual parser patterns, streamlining the setup of single variables, tables, and nested paragraphs.

How it works:

1. Provide a sample CLI output from the command you want to automate.
2. Use guided prompts to generate and refine parser rules.
3. Validate the extracted fields and iterate instantly.
4. Use the structured output in runbooks, assessments, and automation workflows.

Example:

Type show interfaces. Ask for fields like status or CRC errors—AI instantly returns structured data. No regex.

“Get me interface name, status, MTU, and CRC errors.”

AI does the hard part:

It instantly scans the text, finds the patterns, and writes the extraction rules.

You get:

Perfectly structured data in seconds.

The result:

Skip weeks of learning parser coding. The AI builds the automation for you—so your systems can immediately monitor interfaces and trigger alerts.

Zero Regex. Zero Parser Grammar.

You don’t need to write regex. You don’t need to learn parser grammar.

Traditional parsing forces you to become a regex expert just to extract data from CLI outputs. AI Parser Assistant eliminates that barrier entirely.

• No regex writing: Describe what you want in plain English. AI handles the pattern matching.
• No debugging: Regex is notoriously brittle. AI generates validated patterns that work the first time.
• No regex expertise required: Your network team knows the CLI outputs. That’s all they need.

The regex happens behind the scenes. AI translates your prompts into accurate regex patterns, then automatically converts them into standard NetBrain visual parser rules. You never touch the regex unless you want to.

Self-Service Parsing: Turn Any Device Output into Actionable Automation Data

Empower teams to instantly normalize device outputs without relying on specialized parser authors. By enabling consistent, reusable parsing across your organization, you can standardize operational checks, eliminate integration bottlenecks, and rapidly scale automation coverage—all without custom code or vendor dependencies.

Accelerate automation by accelerating parsers

AI Parser Assistant helps teams turn CLI output into automation ready data faster, so more workflows can be automated with less friction and less dependency on niche expertise.