In this video, Sean McDermott of the Find Flow podcast sits down with Ani Gujrathi, chief technical officer for Zenoss, and myself. We dive into the original approach to AIOps, how that has evolved, and how it continues to evolve.
We start by exploring the entire purpose of AIOps — figuring out how to accelerate problem resolution in modern, complex IT environments while dealing with the pervasive problem of monitoring tool silos. Generation 1 AIOps platforms endeavored to do this by collecting events from the plethora of monitoring tools and leveraging machine learning to analyze those events. This essentially amounted to a newer version of event correlation tools, which have existed for decades.
In just a few years, we’ve come a long way from Generation 1 AIOps, and Generation 2 AIOps platforms have made huge strides in providing faster, more accurate insights. They have proven that it is possible to achieve what most believed to be impossible — automating root-cause analysis in massive, complex IT environments.
What will the next iteration of AIOps platforms look like? The speakers discuss how future advances will hinge on trust. When the aspiration (and the necessity going forward) is to automate, there needs to be a level of trust humans have in the machines to perform increasingly complex tasks. How do we achieve this level of trust? As we discuss, some key aspects include what we call bring your own algorithm (BYOA) and explainability. To this end, will the machines eventually replace human beings in IT operations? We talk about why this is highly unlikely and how organizations are simply leveraging AIOps to make better use of human capabilities.
Ani also discusses the three techniques a platform is required to have in order to qualify as an AIOps platform. These are anomaly detection, event correlation and dynamic thresholds. But even among platforms that offer these capabilities, all are not created equal. The key difference among all AIOps platforms is the ability to ingest different types of data for which these techniques can be leveraged.
Check out this episode to understand the past, present and future of AIOps, and how you can use this information to make the best decisions for your organizations.