Management

What Should State and Local Governments Know About AIOps?

Agencies can streamline and automate IT operations by using AIOps tools. Here’s what that could do for governments.

Michael Hickey is an Editor at Manifest.

Artificial intelligence technology has recently caused a stir in the IT world and beyond. But how can state and local governments best harness it? Agencies may benefit from taking a broader view of the current landscape for AI in local government and look at how machine learning assists public organizations, along with AI’s potential pitfalls. But there’s another way states can get a grip on AI: through AIOps.

What Is AIOps?

Artificial intelligence for IT operations (AIOps), as defined by Gartner, “combines big data and machine learning to automate IT operations processes, including event correlation, anomaly detection and causality determination.”

AIOps involves applying AI capabilities such as natural language processing and machine learning models to collect and aggregate the huge volumes of data generated by multiple IT infrastructure components, application demands, performance monitoring tools and service ticketing systems. It helps IT workers separate signal from noise and diagnose the root causes of issues so they can be reported to IT and DevOps teams for rapid response and remediation. The ultimate purpose of AIOps is to automate and streamline operational workflows.

AIOps solutions offer full-stack observability of IT operations, meaning they can manage and address not just security issues but also operational hiccups — such as slow database speeds or lagging application performance — across the board. Issues related to network monitoring, applications, performance monitoring, databases and digital experience monitoring can be addressed automatically with AIOps.

Click the banner below for more content on how IT modernization improves citizen services.

What Are the Benefits of AIOps?

IT operations have only become more complex and chaotic over time with cloud migrations, the exploding number of workloads and applications, and workloads moving from one cloud to another. AIOps can help suppress this noise, which would otherwise be overwhelming to IT teams and contribute to alert fatigue.

“AIOps is an incredibly powerful use of AI and machine learning that allows IT professionals, DevOps professionals and help desk agents to do more with less,” says Cullen Childress, group vice president of product at SolarWinds.

The automation capabilities of AIOps help organizations respond more quickly and effectively to any outages, anomalies or security events. Automated detection and response also lowers operational costs. Plus, its reliance on machine learning means it will consistently improve and be able to predict security events before they even happen, giving IT teams the opportunity to be proactive in mitigating operational issues.

“Without AIOps, the alert fatigue and all of this noise is so loud that you really can’t focus on the signal,” Childress says. “With AIOps, those highly paid IT resources can focus on real problems and accelerate the discovery of an issue. That means organizations don’t need to hire more IT professionals to chase these ghosts.”

Scott Pross, vice president of technology at Monalytic, a SolarWinds company, says that he talks to state and local government IT professionals every day. “I hear the same thing: ‘I don’t have the manpower. I don’t have the time to train them. And I don’t have the resources available.’ AIOps makes it so the system can handle that,” he says.

According to Palo Alto Networks, the core purpose of AIOps platforms is to be able to use telemetry — centralizing data from multiple sources — to prevent and detect issues, then use machine learning to adapt and refine the process.

What Can State and Local Governments Use AIOps for?

There are a number of use cases for AIOps including anomaly detection, alert correlation and clustering, intelligent alerting and network monitoring. Using AI and machine learning, a software provider could develop solutions for monitoring infrastructure and reducing the number of alerts created. AIOps software can analyze anomalies in a way that won’t set off an alert unless it determines the anomaly is a threat or follows a unique pattern.

AIOps is also used for to automate and improve remediation. Once an issue is discovered and elevated, AIOps software ensures that all traces, logs, metrics, entity relationships and dependencies are appended to the incident. With that data, and with data from previous incidents, AIOps can automatically identify the root cause of the issue.

AIOps tools do this by taking data and feeding it into models that can be trained to recognize what a particular environment’s normal state is. AIOps software monitors an organization’s full-stack infrastructure, databases, applications, user experiences, logins and other data to detect when something out of the ordinary is happening.

“The system is trying to find individual data points that share qualities and then match that against what it has seen in the past,” Pross says.

With natural language processing capabilities, agencies can also use AIOps tools to improve IT service management by handling help desk requests. As IT service requests become more complex, organizations can either hire more IT staff to keep up, or they can leverage AI, machine learning and natural language processing via ChatOps (chatbots and virtual agents) to assist users automatically with low-level tickets that don’t need to be escalated to human agents.

With AIOps, your highly paid IT resources can focus on the real problems.”

Cullen Childress Group Vice President of Product, SolarWinds

What Is Domain-Agnostic vs. Domain-Centric AIOps?

Domain-centric AIOps involves applying AIOps to solve specific problems within a certain domain. An organization employing this strategy would use AIOps only for a specific use case such as network monitoring or application performance management. The advantage here is that domain-centric tools are able to provide more pointed solutions to a specific problem than domain-agnostic tools can. For example, a domain-centric AIOps solution would employ an application performance monitoring tool, which has more features dedicated specifically to network monitoring than a domain-agnostic solution has.

“We have a fairly domain-centric approach to the use of AIOps inside our solutions that are built to solve very specific problems as it relates to use cases for database performance, database observability and application performance management,” Childress says. “From a domain-agnostic standpoint, it could be a multitude of different applications that need some type of anomaly detection capability or forecasting capability.”

DISCOVER: 5 questions state and local agencies can ask about AI.

Meanwhile, domain-agnostic AIOps models are general-purpose tools that can solve problems in multiple domains across an IT environment. Domain-agnostic solutions can help IT teams manage all aspects of their IT operations, albeit with less specificity in any particular domain than a domain-centric solution would offer.

Implementing the right AIOps solutions is a matter of determining your organization’s specific needs. As IBM says, the journey is different for every organization, but IT leaders should ensure that the AIOps tools they use have features pertaining to observability, predictive analytics and proactive response.”

What Are Common Misconceptions About AIOps?

With IT automation, there’s speculation that IT professionals will eventually become obsolete. That isn’t a concern, says Childress, and what AIOps really does is free up agents to focus on more important and complex tasks while routine operations are handled automatically. And when new issues arise, it’s IT professionals who need to analyze the situation. AIOps also affords IT teams the time to think more strategically in other areas, such as determining which applications they should build or adopt to drive growth. It’s the position of experts that there’s no substitute for the intelligence and adaptability of human beings.

“AI is hugely valuable, and we’re certainly excited about its uses. Is it going to replace IT? No,” Childress says. “There will always be new issues that come up that we haven’t seen before, that you have to unpack.”

your_photo/Getty Images