AI workbench
SSH, Telnet, and SNMP collection keeps device facts beside AI analysis and its source evidence.
Who it is for
- Network engineers managing devices, topology, inspections, and change evidence
- Operations teams requiring local deployment and auditable workflows
What problem it solves
Network troubleshooting often relies on personal experience and jumps across devices, topology, knowledge bases, and screenshots, making evidence hard to review.
Core capabilities
Topology management
LLDP/CDP neighbor data and NetBox Cable records are used to build and validate topology.
IP endpoint tracing
ARP, MAC, VLAN, interface, and device evidence is correlated to trace endpoint access paths.
AI inspection
Inspection tasks collect device state and combine it with the knowledge base to produce reviewable conclusions.
Batch deployment
Commands or configurations can be sent to multiple connected devices while preserving execution results for verification.
Typical workflows
- Maintain device and address data
- Collect configuration, interface, log, and neighbor evidence
- Run topology, endpoint tracing, or AI inspection tasks
- Review recommendations before batch changes
- Export reports and retain knowledge
Use cases
- Network diagnosis
- Device health inspection
- IP management
- Pre-change checks
Inputs
- Device inventory and connection details
- SSH, Telnet, and SNMP evidence
- NetBox or phpIPAM data
- Local operations documents
Outputs
- Layered network topology
- Endpoint access path
- Inspection reports and health recommendations
- Batch execution logs and audit records
Deployment and data boundary
- The desktop application runs locally and stores credentials and collected data in the user's environment
- AI capabilities require a user-configured model connection
- Production or organization-wide use requires prior licensing confirmation
System requirements
- Windows: 64-bit Windows; exact compatible versions are still being confirmed
- macOS: Apple Silicon M-series chips; Intel Macs are not supported; the minimum macOS version is still being confirmed
Current version and update
1.2.1
Published：2026-07-08
Content updated：2026-07-14
Downloads and verification
Windows x64
- Version
- 1.2.1
- Platform
- Windows
- Architecture
- x64
- Compatibility
- x64 for 64-bit Windows; exact compatible versions are still being confirmed
- Size
- 10.28 MB
- Published
- 2026-07-08
- SHA256
08a1a82e2c5882e987627e889ac8e05aa73560e12f08f7e33dd77c2fc735050a
macOS Apple Silicon
- Version
- 1.2.1
- Platform
- macOS
- Architecture
- Apple Silicon
- Compatibility
- Apple Silicon for M-series chips; Intel Macs are not supported; the minimum macOS version is still being confirmed
- Size
- 16.08 MB
- Published
- 2026-07-08
- SHA256
248d15aaa876b3440b8fd2203dbafa759c9b5201abea3a930fbba4c571e844aa
Evaluation Edition and Extended Licensing
For personal learning, technical validation, small-scale testing, and non-production evaluation.
Existing data remains available in the Evaluation Edition; advanced capabilities follow the extended licensing scope.
Production or organization-wide use requires prior confirmation of the licensing scope.
Evaluation Edition Scope
- Devices: 20
- AI inspections / month: 10
- Endpoint traces / month: 5
- Linux servers: 3
- Batch devices / run: 3
- Audit retention: 7 days
Evaluation Edition Capabilities
- Device inventory, connection settings, and basic collection
- SSH terminal, SNMP collection, and core knowledge-base capabilities
- Basic link modeling and topology
- Local IPAM management and queries
- Local users, roles, recovery codes, and 7-day audit logs
Extended Licensing
- Telnet remote terminal and SNMP templates
- One-click device backup, inspection, and log analysis
- LLDP/CDP automatic link discovery and NetBox Cable synchronization
- Writing collected data to IPAM, data import/export, and migration backup
- Enterprise controller, child-client approval, license reporting, and sync policies
- OEM branding, installer name, icons, and copy customization
When extended licensing applies
- Enterprise production environments
- Larger-scale device management
- Higher-frequency inspection and endpoint tracing
- Longer audit retention
- Organization-wide use
- Custom deployment, adaptation, and technical support
- OEM or brand customization
Known limitations
- macOS supports Apple Silicon only; Intel Macs are not supported
- Minimum Windows and macOS versions are still being confirmed
- AI features depend on a user-configured model service
Frequently asked questions
Does Network AI Ops require a public cloud?
No. Core device, topology, collection, and audit workflows run locally; AI features use the model connection configured by the user.
Which installers are available?
Version 1.2.1 provides Windows x64 and macOS Apple Silicon installers. Intel Macs are not supported.
What does endpoint tracing produce?
It combines IP, MAC, VLAN, gateway, aggregation, and access evidence to identify the access path and final switch port.
Can the Evaluation Edition be used in production?
It is intended for learning, technical validation, small tests, and non-production evaluation. Production or organization-wide use requires licensing confirmation.
Product access
Product screenshots
Video demos
AI Workbench
Natural-language tasks, knowledge selection, run traces, and diagnosis output in one workspace.
Topology Management
Layered topology, device placement, links, and SNMP status in one view.
IP Address Management
Manage VLANs, prefixes, IPs, scanning, and endpoint trace tasks.
Device Management
Maintain device inventory, connection profiles, collection history, and config versions.
Batch Deploy
Review target devices, scripts, uploads, execution logs, and audit results together.
Knowledge Base
Sync Markdown, PDF, Office, and CSV documents into the diagnosis flow.
Settings Center
Configure models, embeddings, SMTP/Webhook, backups, and enterprise controls.