AIOps: The riddle wrapped in the mystery inside the enigma

It didn’t take long into my tenure with Forrester to realize that there was confusion in the industry regarding AIOps. Both technology vendors and industry companies openly asked me what the term meant, because they are hearing many variations. They gave me examples of some as simple as a chatbot and others as complex as end-to-end dedicated platforms. We won’t ultimately answer the question in this blog, but we can work toward that end. My discussion with these groups brings me to you. How do you view it? 

  1. Is AIOps a technology, a platform, a solution, a capability, or a combination of each? 

  2. Does one size fit all? 
  • Is the end state the same for everyone? 
  • Will every effort look identical in how it achieves its end state? 

  • Are there commonalities across industries and organizations? (tools, skills, approaches, objectives) 

I recognize that these are loaded questions, but we need to face them. We need to come to some agreement on them, and then we can move forward. Ultimately, we need to arrive at a common understanding that clears up some of the confusion around the term AIOps. We need to agree on some fundamental principles about the concept of AIOps. 

The first step is to provide some framing around the words in the first question to ensure that we’re on the same page. These are not formal Webster’s Dictionary definitions, so work with me on this. 

Solution Vs. Capability 

  • A solution is about the end-state consumer/customer. This is your “why.” It’s why you’re doing something — why it matters. 

  • A capability is about you. This is your “what.” It’s what you need to do to deliver the solution. 

Technology Vs. Platform 

  • A technology implies a component more singular in nature that is installed and delivers on an end objective. 

  • A platform — on the other hand — is typically comprised of multiple technologies that collectively deliver on a broader, multifaceted objective. 

Now The $1M Question: “What Is AIOps?” 

It should be obvious: It must be a combination of all the elements described. They all must be contributing to the overarching purpose of this concept we’ve been calling AIOps. Organizations are not homogeneous entities, though. They have different needs at different points along their maturity path. That means the elements described above are needed but in varying degrees, in different quantities, at different points in time for each organization. All four elements play a role in completing the AIOps story — the challenge is answering: How does each contribute, and how do we know how to put it together? 

The first two terms (solution and capability) are the easiest ones to understand, because to achieve a solution, you need a capability. To be more exact, you need lots of capabilities put together specifically for you to deliver the solution. Some capabilities you may already possess; some you need to acquire. 

The second two terms (technology and platform) are trickier. This is mainly because a platform will have one or more technologies. A discrete technology may or may not be part of a formal platform. In summary, a technology could itself be viewed as a platform depending on the granularity in which it’s defined, whereas some technologies are simply point solutions, independent of a platform. 

Think of it in terms of automobile manufacturing. An automobile has multiple assemblies — or what we’ve been calling technologies. These assemblies — windshield, frame, drivetrain, engine, dashboard, etc. — are put together to produce the automobile as we know it. Each assembly, however, can be made up of other, smaller assemblies, or subassemblies. Each element within an assembly, no matter how small or large, simple or complex, contributes to the final product. No element can be devalued, because they all play valuable and vital roles. 

Similarly, we must determine the various elements that contribute to this concept we’ve been calling AIOps. We need to leverage it to move from technology operations to modern technology operations management or, better yet, automated technology operations management. Before we do this, though, you also need to answer: Does one size fit all? 

Does One Size Fit All? 

If you didn’t see this coming, then you haven’t been paying attention. Of course, one size does not fit all. A regional bank has different needs than that of a global industrial equipment manufacturer. Even two regional banks are likely to have different capabilities, skills, and toolsets. This uniqueness is what will impact the approach each takes to meet the business needs. Insight into everything from the infrastructure out to the user experience, and everything in between, is what industries and organizations commonly desire. The purpose of having this insight is to act on it in a proactive — or, better, predictive — manner using automation. 

I recognize that the primary question — “What is AIOps?” — is still not truly answered in this post. I hope, however, that this blog has pushed your AIOps conversation away from it simply being installed software and instead toward it being a comprehensive, operational concept where insights-driven automated actions are the goal. 

The problem is that few have the insights to act on — never mind possess — the ability to automate actions based on it. You need this, even though most groups are not sure how to make it happen.  

This post was written by Principal Analyst Carlos Casanova and it originally appeared here.

Source link

3 thoughts on “AIOps: The riddle wrapped in the mystery inside the enigma

Comments are closed.