3 minute read time

3 tips for putting AIOps into action

by   in IT Operations Cloud

Top tips for putting AIOps into action

IT teams are constantly challenged by rapidly growing data volumes generated by IT infrastructure and applications. Their job is to sift through this never-ending data and produce actionable insights. But the reality is that IT teams often work in silos, resulting in disjointed work and longer than needed MTTR. To address the problem, organizations turning to AIOps.

Here are my top three tips for putting AIOps to work. I’ll provide real-world examples along the way:

1. Unify your monitoring into a single tool

Most organization’s applications are sprawled across different environments, so it’s getting increasingly difficult to find and fix a problem when it inevitably arises. That’s what brings me to the first tip—to unify your monitoring into a single tool, because you can’t manage what you can’t monitor.

Consider, for example, a large telecommunication provider with thousands of applications sprawled across five data centers. As you can imagine, these applications contain vast amounts of data that needs to be monitored. That’s why the company saw the value of putting all this data into one monitoring tool. A central management structure creates a single pane of glass where each item in the data center can be traced and managed through metrics, so they never miss a problem.

See how Turk Telekom unified monitoring

2. Tie IT to business impact

There was a time when siloed consoles were used to alert outages. The consequence was that the team expended effort to understand the culprit—was it the firewall, a database, or applications? This fragmented monitoring model highlights the misalignment between IT and the business impact. But how do you solve this problem?

One of our customers had this same problem and they started by improving their dashboards. Dashboards give real-time insights into IT and business status. If dashboards don’t solve the problem, topology maps help with prioritizing incidents based on business impact. For example, you’d rather fix a revenue-generating application as opposed to an internal application outage. The diagram below, with the application and all its components, shows where to start solving the problem.


Topology or dependency map showing which devices and services are part of the application.

Learn how Vodafone uses dashboards to align IT closer to the business and reduce alarm reduction by over 70%.

3.  Automate, automate, automate

The biggest AIOps tip is centered on automation. That’s not surprising since all of our heterogeneous systems with all their complexity demand it.

Thanks to automation, one of our customers fixes 95% of known problems with no human intervention—saving about $4M annually. Even if you’re unwilling to take a hands-off approach, you can realize big benefits. For example, by automating runbook solution, you can save typing time and, more importantly, typing errors. Many major outages are due to the wrong command being typed, one customer told me.

Observability with AIOps

Our complex systems demand more time than we have to give, even with AIOps in place. That’s where observability comes in. But how does observability relate to AIOps?

Observability is a capability. An AIOps tool with observability lets you see what’s happening inside complex systems. You need it because the complexity and volume of data are too high to handle manually. AIOps capitalizes on observability, providing automation to help fix problems once they are found.

Turn on AIOps and you’ll strengthen the reliability and performance of your entire IT estate.

 Related content


Have technical questions about Operations Bridge? Visit the OpsBridge User Discussion Forum.

Keep up with the latest Tips & Info about Operations Bridge.

Do you have an idea or Product Enhancement Request about Operations Bridge? Submit it in the Ops Bridge Idea Exchange.


Operations Bridge