TL;DR: HPE is bringing AIOps into network operations through HPE Aruba Networking Central and HPE Juniper Networking Mist and Marvis. These capabilities can help teams reduce manual troubleshooting, accelerate root-cause analysis, automate selected remediation, and extend the capacity of existing network staff while keeping engineers in control where human judgment is still needed.
IT teams have been dealing with talent shortages for years, and the gap isn’t getting any easier to close. That’s particularly true in network operations, where the percentage of organizations reporting difficulty hiring and retaining people with networking expertise jumped from 41% in 2024 to 52% in 2026.
At the same time, much of that hard-to-find expertise is absorbed by day-to-day operations. It’s necessary work, but it leaves less time for staff to use their expertise on problems where their experience matters most.
AIOps offers an opportunity to shift that balance. Used well, AI and automation can reduce some of the operational burden and help teams make better use of the expertise they already have.
The question for IT leaders is where that shift makes sense. They need to ask:
The challenge isn't only finding enough network engineers. The skill set those engineers need is also expanding as networks span cloud, security, automation, AI, and increasingly distributed infrastructure. At the same time, troubleshooting widely distributed networks requires engineers to piece together information across wireless, switching, WAN, cloud, applications, and other parts of the environment.
The problem really has become simple: network teams are being asked to know more and investigate across more of the environment, often without a corresponding increase in staff. That makes relying on manual operations increasingly difficult to scale.
Network professionals already spend an average of 29% of their workday troubleshooting and navigating alerts from different systems and tools. And with the massive amounts of alerts coming in from different tools in a single day, significant time can be spent determining which problems require attention before getting to the work of diagnosing them.
AIOps can help reduce some of that manual workload. In one study of network security operations, researchers found that after introducing automation, 44.9% of alerts could be handled without human involvement. While not every task can or should be automated, AIOps can shift where engineers need to enter the troubleshooting process.
That changes where the engineer enters the process, too. Instead of starting with “Where should I look?”, the engineer can start with, “Does this diagnosis make sense, and what should we do about it?”
For teams already struggling to find experienced networking talent, that shift can make a meaningful difference. AIOps won’t create more network engineers, but it can help existing teams extend their capacity and manage increasingly complex environments without scaling manual effort at the same rate.
Let’s look at some areas where AIOps can add real value, helping network teams move more quickly from detecting a problem to investigating, remediating, and validating it. AIOps can:
When an issue surfaces, engineers may need to determine which alerts are related, gather data from multiple sources, and isolate the root cause before they can even begin resolving it. Research into AI-assisted network management shows that AI can be especially useful for network engineers in identifying and diagnosing issues across systems. Some commercial AIOps platforms are already putting this approach into practice. HPE’s Marvis AI Assistant, for example, uses network telemetry and other operational data to help identify likely causes and point engineers toward where to investigate.
Knowing where to look is often part of the challenge. Conversational AI gives engineers another way to interrogate operational data, allowing them to ask natural-language questions rather than starting with the right dashboard, query, or CLI command. That can be useful when an issue crosses domains or when the person investigating doesn’t yet know which dataset will contain the answer. It doesn’t eliminate the need to understand the network, but it can reduce some of the work required to get to the information an engineer needs.
Configuration management is another area where researchers are exploring AI-assisted operations. Current approaches can analyze configuration histories, look for inconsistencies, compare conditions before and after changes, and help verify network configurations. That could give teams another layer of validation around changes across increasingly heterogeneous environments. However, this is also an area where human oversight matters: AI-generated configurations need to be validated before deployment, since an incorrect command or configuration can create the very outage the team is trying to avoid.
AIOps can also use historical and real-time network data to identify anomalies and trends that may indicate deteriorating performance, capacity constraints, or potential failures. While the goal isn’t to predict every outage, the tools here can give network teams more opportunities to investigate developing conditions before they become user-impacting incidents. Proactive assurance, anomaly detection, predictive maintenance, and capacity planning are all being explored as AI applications in network operations.
AI-assisted remediation doesn't have to mean handing network control over to an autonomous system. While 87% of network professionals say they prefer AI-powered tools for remediation and optimization, 41% prefer a human-in-the-loop approach where AI guides the action without executing it. That leaves room for a spectrum of approaches, from recommendations to automated remediation. HPE's Marvis Actions is one example: teams can keep administrators in the loop for recommended actions or authorize self-driving remediation for specific supported scenarios.
AIOps is also becoming less of a standalone tool and more of a capability built into the platforms teams already use to operate their networks. That means organizations refreshing wireless, switching, WAN, security, or cloud-managed networking have an opportunity to evaluate operational capabilities alongside traditional requirements such as performance, reliability, and security.
That's the approach HPE is taking across its networking portfolio. AI-native operations span HPE Aruba Networking Central and HPE Juniper Networking's Mist and Marvis capabilities, giving organizations options to apply AI-assisted operations across wired, wireless, WAN, and other network environments.
The right strategy depends on your environment, your team, and where manual operations are currently creating the greatest burden. The key is to start with the operational problem (not the AI tools!) and evaluate solutions based on whether they actually make the network easier to manage.
Not sure where AIOps fits into your networking strategy? VLCM can help you evaluate your current operations, compare AIOps options, and identify opportunities for AI and automation across your network.