On the device
What the agent checks, what it may fix, and what it never touches.
The agent is also the RMM agent, so the AI reads a device the way your technicians do, and acts only through the automations you enabled for that client.
30 days · No card required · Work account
AI safety and security
The AI asks before it acts.
1. It reads
The ticket, the device and that client’s knowledge.
2. It asks
The requester, your technicians or the site manager, as you set per fix.
3. It records
The run, the approval and the result, on the ticket or in History.
What it reads
Health, hardware and security status, installed apps and versions, patch status, services and processes, disk, network reachability and the device clock, plus recent events and logs when a ticket needs them. Microsoft 365 and Google Workspace checks on a ticket are read-only.
An AI RMM that asks before it touches a device →What it may fix, with your approval
Install or update an approved app from your catalog, restart a stopped service and end a runaway process, free space on a full disk (temporary files only, never personal files), correct the clock, turn off a broken browser proxy, find the firewall rule blocking a service, and run your reviewed PowerShell or Bash scripts on one device or many.
Automation that runs only what you allowed →What it never does
It changes a device only through a fix you enabled, with the approval you set for it, never runs an automation you did not enable for that client, and never runs a script the AI drafted until a technician has reviewed it. Patching follows your patch policy; the agent does not decide what to install.
Permissions, approvals and the record →In our recorded demonstration
A VPN ticket: the clock was the cause.
A user could not connect to the VPN. The agent found the laptop clock seven minutes ahead, asked IT to approve a correction, corrected it and asked the user to try again. The ticket kept the finding, the approval and the result.
See the demonstration →Questions MSPs ask
Windows, macOS and Linux, with available actions depending on the platform. The agent connects out over HTTPS only and uninstalls like any app.
Yes, as devices under the client, with the same approvals and the same record.
The agent and every tool it runs are signed by NetZen AI Inc, and it checks each script’s signature before running it. For strict EDR products we publish the folders and processes to exclude.
From an MSP using NetZen
“NetZen seems much more aligned with where support operations are heading rather than just adding an AI integration onto an older model.”
Vik Joshi, Cinema Hero Solutions
Read the customer story →Try the NetZen platform
Thanksgiving Offer
$50$100USD per technician per month*
Connect your own computer, send the sample app request and follow the approval and the device action.
Continue exploring
Control