E-Handbook: How to get network data analytics that really helps your team Article 4 of 4
Tip

Using AI, machine learning in networking to improve analytics

The barrage of alerts modern network analytics tools create is problematic, but the capabilities advanced tech like AI provides can help ease the burden to network operations.

Anyone who decides upon a career in network operations likely expects a life of adventure and regular testing of one's problem-solving skills, but often they don't make a deep consideration of how all of those adventures and challenges will play out in the nights and weekends ahead. However, that consideration is essential, given the sheer volume of traffic, the scope and scale of modern organizational networks and the always-on and mission-critical nature of network operations.

Fortunately, there is help in the form of modern network analytics tools. These applications -- whether cloud-based or installed on site -- are a resource that no network today should be without. Such tools -- anchored by processes that can rip through vast volumes of operational data faster and more accurately than any mere mortal could -- assist operations pros with real-time insights that would otherwise be unavailable. At the same time, network operations analytics is now getting much better with the application of AI and machine learning in networking analytics tools in the operations arsenals. These advanced technologies complete the feedback loop between analytics and management consoles -- and represent the best hope yet for freeing up nights and weekends of network operations personnel.

AI has been with us for decades, and the concept is quite simple: Apply the knowledge of human experts and embed it in algorithms that run in ever-increasingly powerful and cost-effective processors and storage systems. Machine learning enables the performance of those algorithms to improve over time. Both AI and machine learning in networking are increasingly available as cloud services, which increases their cost-effectiveness and eliminates the need for costly servers and associated equipment to be maintained on site.

When AI and machine learning met NetOps

I like to look at AI, ultimately, as the magic hand on the management console, tweaking settings on a 24/7 year-round basis.

So, how to use AI and machine learning in networking operations? Analytics is just part of the equation. Even with the insights these network analysis tools produce, it's not often clear what steps should be taken next to remediate a problem or to prevent such a problem from occurring again. With AI part of the mix, these applications can not only determine the root cause of an issue, but automatically implement the correct fix. I like to look at AI, ultimately, as the magic hand on the management console, tweaking settings on a 24/7 year-round basis in response to what's been identified in analytics and considering perhaps hundreds of variables, traffic patterns over time, local policies and more.

Here's where machine learning in networking comes into play: As optimal solutions to identified problems are proven safe and effective, the AI-enabled network analysis tool integrates this knowledge just as a human operator would. So the tool gets better, faster and thus more productive. While we can't expect perfection here, just as we can't from humans, AI and machine learning get us a long way down the road to cost-effective optimization, including optimal productivity for operations pros and end users alike.

What will network ops pros do in an AI-fueled world?

Keep in mind that AI and machine learning in networking analytics tools perform essentially the same set of functions as operations professionals always have -- just faster, more accurately and without the need for downtime.

And what will operations experts do with their newly discovered free time? Well, sleep and a nice vacation every now and then come to mind, but it's likely many will focus on network planning, testing new technologies and tools, working on optimizing usage policies, helping users get the most from the network, and the dozens of other tasks where human knowledge, experience and reasoning are essential.

Next Steps

What an AI-driven network looks like

Dig Deeper on Network management and monitoring

Unified Communications
Mobile Computing
Data Center
ITChannel
Close