Automation & Runbooks
Let routine problems fix themselves, keep a person in the loop for everything else, and see every fix on the record.
Runbooks: fix it, check it, page only if it failed New
A runbook says: when this happens, run these steps in order, check whether the problem cleared, and page a person only if it did not.
Chain the steps
Each step is one of your approved scripts, with the device’s drive, service or certificate filled in. After the last step TechSavve checks whether the alert cleared, and pages a person only if it did not.
Four fixes ready to add
A shared library covers a full disk, a stopped service, a stuck print spooler and an expired certificate (renewed with certbot or win-acme). Each is added switched off.
Safe by design
Nothing is switched on until you say where it applies. Warning alerts never page, down devices are never touched, and a daily cap limits how many fixes may run.
On the record
Every run shows its steps on the alert, the incident and the ticket, and counts in the customer’s business review as a problem solved before they noticed.
New. Runbooks were tested end to end with simulated certificate tools and in our own demo, and have not yet run on a customer’s devices. Start with one customer and one runbook.
A script library you can schedule
Nearly 200 built-in scripts, and your own beside them: 111 PowerShell, 71 shell and 14 Ansible playbook checks that run from Linux collectors.
What is in it
Diagnostics, storage, networking, security, services, updates and reboots, Active Directory, SQL Server and other databases, web servers, containers, VMware, Hyper-V, users and groups, Microsoft 365, and VPN and identity.
Run on a schedule
Run a script daily, weekly or monthly across a customer, a site, devices with a tag, or a mix of customers, with a preview of what will be hit first.
Results can alert
A scheduled script that keeps failing raises an alert and, past your threshold, a device incident, through the same tickets and on-call as everything else.
Credentials stay in the vault
A script asks for a credential by name. The collector fills it in when the script runs and masks it in the output.
Approvals
Hold scripts for an administrator’s approval before they run. A request that is not answered expires after a day.
Every version kept
Edit a script and the earlier versions stay, with a side-by-side compare and a restore button.
Ask the NOC New
An optional assistant for your technicians, on your own Anthropic API key, so you pay Anthropic per use and there is no fee from us.
What’s wrong?
On an alert, an incident or a device, one click explains it in plain words from the facts TechSavve already holds.
Draft the ticket
A reply, a summary or a resolution note, drafted for a person to edit and send.
Ask the NOC
A chat that answers questions about your customers, devices, alerts and tickets using read-only tools, and sees only what the person asking is allowed to see.
Under your control
Off until you add a key. Choose the model, set a monthly limit, and switch each feature on or off. Credentials, vault entries and monitor settings are never sent, and calls are logged by size, not content.
New. The assistant was tested with a simulated provider and has not yet been run against the live Anthropic service by a customer. Add a key and try it on your own data.
More that runs by itself
Quieter alerts
One-click suggestions that wait, raise a threshold, mute a backup window or learn what is normal. Read more.
Checks for the rest of the stack
NAS and backup systems, UPS units, printers, Redfish servers, firewalls, Tailscale and Kubernetes. Read more.
Auto-remediation
Run one approved script when a monitor keeps failing. Read more.
See TechSavve RMM in action
Book a free demo with the DFWTechPros team.