Welcome back! In the previous blog, we discussed what AIOps is and the significant role it plays in ITOps with our panelists: Carlos Casanova, Principal Analyst at Forrester, and Gowrisankar Chinnayan, head of product management at ManageEngine.
In this blog, I am going to walk you through the rest of our discussion. Our talk revolved around the reasons why organizations should adopt AIOps.
In today’s world, businesses have come to rely on digital technology, which in turn gives birth to complex digital problems. The diverse, ever-changing IT landscape ultimately pushes organizations and IT professionals to lean towards AI and ML to ensure smooth functioning of the entire IT environment, which ultimately leads to smooth functioning of the organization.
Casanova stated that in reality, almost 70-80% of organizations had started investing in insight-type technologies like AI and ML. He reckoned these tools enabled a certain simplicity in the environment because they could ingest data from a whole host of sources and coalesce it all into one bundle, presenting the IT teams with a composite view of their IT environment.
Another reason Casanova stood by AIOps was because the mundane and repetitive tasks performed by workers on a regular basis could be taken off their plates, which allowed them to focus on work and be productive. This would ultimately increase the overall efficiency of their organization.
After listing the reasons to adopt AIOps, we gradually moved on to the challenges organizations might face.
1. Insufficient skill sets
Nowadays, a lot of simple tasks are handled by technology. But before you tell your employees to hand over these simple tasks and perform sophisticated ones, you should ask yourself these three questions:
Are our people skilled enough to do that?
Do we need to provide them with training?
Is it the right time to take away the simple tasks from them?
One way to avoid this hurdle is to ensure that you have a strong learning culture in your company.
2. Different capability areas
When dealing with capability areas, it is important for an organization to evaluate its maturity level and how much further it is willing to invest in its maturity beforehand. Casanova had released a reference architecture to point out the 18 different capability areas on which organizations needed to focus.
3. A lack of trust in AI
Change is not easy. It’s as simple as that. Many organizations face the tiresome task of making clients realize that AI is not going to take over the world and that it’s just a tool that enhances IT operations. Even IT professionals on the front lines are skeptical about AI, all the while knowing that they are at the peak of another digital evolution.
4. A lack of awareness about data-driven technology
The first step in AIOps predictive analysis is to compile data. Despite understanding the importance of this step, most organizations fail to act on it. Some organizations that were once considered exclusively brick-and-mortar took the challenge of the pandemic head-on and spectacularly made their presence known. Interestingly enough, the companies that struggled the most were the ones that weren’t entirely sure whether to invest in this technology.
Then, I asked Chinnayan how ManageEngine was going to handle these challenges as a vendor.
Chinnayan answered that AIOps had to become a commodity rather than a luxury.
Until a few years ago, AIOps was offered by only a few IT companies. It required a massive investment and wasn’t affordable for all. So Chinnayan suggested that the ideal option was to opt for solutions. He felt it was wise to seek out the products and attempt to utilize AIOps to handle simple use cases and problems before adding more capabilities.
Chinnayan stated to do what fit an organization the best at that moment in time. And if that is all they could afford that day, they could revisit it after a few years and invest more on the back end for monitoring and whatnot.
I completely agreed with his statement.
Then, I posed the most pressing questions to the panelists:
How do we establish trust in AI among people?
How do we address the memes surrounding the capabilities of AIOps?
How do we help bring transparency?
Every now and then, a lot of memes about AIOps pop up. The ability of AIOps to predict values based on tremendous amounts of data and the time taken to accomplish this task are constantly questioned.
Chinnayan expressed that one of the major issues causing distrust was lack of context. Just ingesting all the data, events, and alerts into the system wouldn’t allow the AIOps tool to give proper suggestions and insights because the system may not know the context. To avoid such a situation, he recommended using discovery and topology mapping.
Casanova agreed and said context was really the secret sauce. Chinnayan also felt that involving humans in functionalities like workflows would help promote the trust factor further.
That’s it for part two of my summarization. In the next and final part, we will discuss the future of AIOps in the IT space. Stay tuned.
To learn more about our solutions, click here.
