// Managed IT

What managed IT monitoring should include

A dashboard is not a service. Here is the scope, alert routine, and handover to look for before you pay for managed monitoring.

Managed IT monitoring is not a dashboard you log into when something feels wrong. It is an agreement that someone owns the signals: defines what normal looks like, watches the alerts, decides what needs action, and reports what changed. Monitoring software is a tool. Managed monitoring is the operating routine around that tool, and the difference is usually visible on the first bad night.

Two proposals can both mention Zabbix and describe different services. One installs an agent and gives you a login. The other sets thresholds for your operation, routes critical alerts to a person, documents what was remediated, and reviews the setup when infrastructure changes. Only the second is managed.

What the service should cover

Start with the systems whose failure stops work, not with every metric the tool can collect:

The inventory should be written down. “We monitor the infrastructure” is not a scope; a list of devices, sites, and services is. It also makes additions explicit when the environment changes.

Alerts should produce a decision, not a notification

Every alert needs a severity, an owner, and an expected action. If the recipient cannot tell whether to restart a service, open a ticket, or wake someone up, the alert is incomplete.

Thresholds should come from the operation, not the tool’s defaults. A file server and a public website do not share a definition of a slow disk. The same applies to business hours: an alert at 10 AM and an alert at 2 AM may need different routing.

The published plans route alerts by email and Telegram. Ask any provider: who receives the alert at 2 AM, what are they allowed to do, and when do they escalate? If nobody can answer, the service is a dashboard with an invoice.

Where AI fits, and where it does not

AI is useful when it reduces the work between an alert and a decision: correlating events, summarizing telemetry, or handling a narrow, repeatable fix. It does not replace a tested restore, a network diagram, or someone who understands business impact.

AdvisoryProTech runs three custom AI operators alongside the monitoring stack. Vali compiles daily reports, aggregates Zabbix telemetry, and sends summaries to Telegram or email. Foso audits endpoints and firewalls, flags unsanctioned access, missed patches, and malware patterns, and can block offending IPs. Pinky handles narrow self-remediation, such as restarting a stalled web service or clearing a bloated temporary directory before a critical alert.

Automation acts inside a defined boundary: access changes, architecture decisions, and business trade-offs stay with a person. “AI-powered” is not a scope, and it is not a substitute for an owner.

What onboarding should look like

Start with a look at equipment, network, access, and backups: the free diagnostic on the managed IT page. The first pass is to find what can take the business down now: a backup that has never been restored, a full disk, a server without updates, shared passwords.

Then stabilize the immediate risks, turn on monitoring, move requests into the helpdesk, and agree hours and people. Ongoing work begins after that: alerts, tickets, and a periodic review. Monitoring a broken baseline mainly produces accurate alerts about a broken baseline.

What the published plans include

The monthly plans are published in USD on the pricing section:

There is also a one-time Security & IT Baseline at $990 for a small team that is not ready for a monthly plan. Do not buy a monitoring plan for systems you do not have, and do not assume a one-time assessment watches anything after it ends.

Questions to ask before signing

How to judge whether it is working

Do not judge it by the number of alerts. Judge it by whether the same incident keeps returning, whether alerts lead to action, and whether the team can explain what is covered. A monthly report is useful only if it changes a decision: expand a disk, fix a backup, replace a device, or stop monitoring a service that no longer matters.

Look for gaps after changes. New servers, a moved site, or a replaced firewall can sit outside monitoring for months if nobody owns the inventory. If only the provider can read the setup, you bought a dependency, not a service.

Bottom line

The tool matters; the routine around it matters more.

To compare that routine with your setup, the first 30-minute conversation is free and remote, on US East hours. Bring the systems that would hurt most if they fail. We will scope a monitoring plan or say a one-time baseline is the better first step. The infrastructure and maintenance pages show what sits around monitoring.