Considering that the language that underpins artificial intelligence (AI), LISP, turns just 60 years old this year, there can be little doubt now that the technology itself has been adopted by the masses. AI has become a critical part of how businesses operate in an extremely short stretch of time – in 2017 alone, 61% of business had implemented AI in some way. Meanwhile, worldwide spending on such technology jumped by over 50% last year to a staggering $19 billion.
A natural byproduct of such unprecedented growth and adoption has also been its impact on more traditional approaches, especially to IT, such as operations. This has given rise to “AIOps”, whereby AI is applied to either enhance or replace, partially, a number of IT operational processes. Given this relatively new approach, Gartner already predicts that 25% of enterprises will be using AIOps by the end of this year.
Moving beyond the nitty gritty, AIOps is really seeking to deliver insight into the digital experience, but also why such technology isn’t working, or even worse, breaking down. While this is extremely powerful, there are a number of data visibility issues that affect AIOps, on its own. This is why, in order to address such data gaps, specific measures are taken that compliment AIOps technology.
In the digital era, undoubtedly, the experience is now vital for every company and brand. Yet with this in mind, how can you monitor to protect this experience? Firstly, getting the right data is paramount. In light of this requirement, IT ops should have visibility to address all the elements that affect the digital experience, but also, crucially be able to address the “why” when digital experience breaks.
How can it do this?
Visibility should include digital experience monitoring (DEM) including HTML server availability and response time, page load and web transaction data. Yet more needs to be known, especially in the modern, cloud and Internet-dependent era. For example, a website or a service, such as a Salesforce API endpoint, may not be responding in a timely manner, but what can the ops team actually do with such information? It simply isn’t enough to go on because they still need to know why it is happening in order to actually solve the underlying problem.
As illustrated by this example, we can really see a significant gap in most IT Ops visibility architectures in place. This is because the vast majority of IT Ops visibility is based on passive data collection from different pieces of an overall puzzle it still has control over.
In today’s software-as-a-service (SaaS) environment, IT teams simply don’t have oversight, or even control over the myriads of external apps, services, infrastructure and Internet networks. This will only become a bigger issue as there will be no slow-down in the use of SaaS, as evidenced by Gartner’s estimation that the market will grow by 17%. In reality, what does this actually look like? For example, your IT team can’t input application performance management (APM) code into a SaaS provider’s software, and they also can’t gather infrastructure data from a network that doesn’t belong to you. As Gartner, pithily puts it, “Infrastructure and operations leaders must rely on outside partners.”
Yet without having data on this vast part of the current IT landscape, you’re left with a pretty insurmountable data gap that means that no level of analytical intelligence can actually make up for it.
Undoubtedly AIOps has a role in dealing with IT ops challenges, but it’s not a panacea for all the issues raised. For example, AIOps, at its core, doesn’t address the visibility data gap around understanding the impact of the Internet and other non-IT-controlled assets on the digital experience.