Today’s IT systems are incredibly complex, with distributed architectures, multi-cloud containers, microservices, apps, Big Data, in-memory systems, and real-time streaming. It is almost impossible to keep up with it all, but technologies like Application Performance Monitoring (APM) and Business Process Management (BPM) have evolved from the early generation of IT Operations Management tools into AIOps, which combine big data and ML functionality to enhance and partially replace all primary IT operations functions.
Although 2021 might not be a transformative year for AIOps, in retrospect, it might be seen as the time when promise met reality. The five AIOps trends mentioned here – observability platforms, data agnostic tools, chatbots, hyperautomation, and virtualized operations and service management – are helping IT operations become more informative, productive,
Observability Platforms
Right now, the AIOps solution landscape is highly fragmented, but software vendors and ancillary technology providers are creating what’s known as observability platforms that allow IT operations to get a deep view into their company’s operational data and systems. Technologies such as the cloud, DevOps, apps, microservices, containers and container orchestration are increasing the velocity of data as well as speeding up the process of going from programming code to full production.
Modern observability platforms cut complexity down to size. They select and present relevant insights to users, while also proactively solving potential issues. For example, a front-end monitoring system can potentially identify JavaScript issues that might be about to overwhelm a system. Performance issues can also be looked at prescriptively. For example, potential out of memory issues could be alleviated long before they become problematic. Alerts for front-end problems, which often originate deep in the application stack or even within conflicting infrastructure components, can also be sent out to necessary parties long before any problems arise.
Data Agnostic Tools
Today, IT legacy systems are being augmented with all kinds of new software products, both commercial and open source ones. This creates a hodgepodge of systems that include a dizzying mix of hardware, software, infrastructure tools, and apps that aren’t always capable of communicating with each other.
The new AIOps solutions allow for greater insights into the data. The more varied the data inputs, the more powerful the AIOps algorithms will become, leading to more insightful data coming out of them. The goal is to build an encyclopedic understanding of the system data and processes, so that issues can be proactively addressed.
Chatbots
Chatbots are considered a form of AI, Natural Language Processing, to be exact. They mimic human intelligence by interpreting a user’s query. They can help operation personnel get instant answers to all kinds of technical questions. Simple chatbots are relatively easy to build and they are everywhere, but the second generation of chatbots, which are rolling out now, can use NLP in a highly intelligent way. Once queried, they can set processes in motion that can set up alerts or provide answers to much more complicated questions. They will reply to queries with smart responses, in a natural, human tone.
Chatbots and virtual support assistants (VSAs) can provide automated support, cutting down the need for live customer service agents. AIOps can be used to improve employee productivity by using chatbots to deliver friction-free answers to a list of standard employee questions or problems. NLP can power chatbots to take the load off the service desk, handling basic inquiries into IT issues, like password resets.
Hyperautomation
Listed as Gartner’s number one technology trend for 2020, “Hyperautomation refers to the combination of multiple machine learning, packaged software and automation tools to deliver work.” It also, “refers not only to the breadth of the pallet of tools, but also to all the steps of automation itself (discover, analyze, design, automate, measure, monitor, reassess).” Put simply, hyperautomation deals with the application of advanced technologies to increasingly automate business processes as well as augment humans.
Hyperautomation allows companies to create a digital twin, i.e., a digital replica of a living or non-living physical entity that allows organizations to visualize how its functions, processes, and KPIs interact to create value for that organization. A digital twin provides real-time intelligence that can help a business gain instant business intelligence that will be useful long before it could in the past.
Virtualized Operations and Service Management
This trend dovetails with the current COVID environment in which employees have been forced to work from home, social distance, and cut down on nonessential travel. The work-from-home trend had been picking up in recent years, but it moved into hyperdrive in March 2020 when the pandemic hit with full force. It’s a trend that is probably going to have lasting consequences.
In the future, the majority of an AIOps team might be addressing an incident while working on a virtual Network Operations Center from multiple worldwide locations. Teams of experts could be rapidly deployed to handle a particular problem or a whole host of issues. Then disbanded as quickly once the issue has been resolved.
Conclusion
“The only constant is change,” as the saying goes and AIOps is embracing a continuously evolving model that will proactively help, it not self-heal, IT operations. Although 2021 might not be a transformative year for AIOps, it will take operations management to a new level. It will see the consolidation of some old technology, the embrace of several new ones, as reveal the true power of virtualization for service management
