Start AI With the Problems Slowing IT Down

Hey everybody, we're at the ManageEngine User Conference in Orlando and I'm here with my good friend Rakesh and we're talking about, well, AI and how the IT operations product portfolio is going to evolve because, well, to be fair, the ManageEngine portfolio is pretty rife. Yeah. And so my question to you, I guess to start with- Mm-hmm ...

is how will that all evolve in the age of AI? Because it seems like AI agents are going to be these interfaces to all these different things, and maybe I don't need to spend as much time mastering the GUI for each one or how- Yeah ... is that going to play out?

Yeah, that's an interesting question. And, yes, so we are in the age of AI now and we see a lot of our customers experimenting with AI agents. If you think about it, with the AI agents basically executing your tasks, the UI becomes less and less important, because the AI agents are interacting with the software in the background through APIs.

And these agents are also communicating amongst themselves through agent-to-agent communication. Now, though the UI might sound like it is getting less relevant, the underlying infrastructure that's powering these AI agents with the data, that's where ManageEngine is focusing on a lot. Because for these AI agents to go in and execute something that you're asking for, you could probably just be asking, "Okay, tell me what happened in the last 12 hours in my IT environment.

What are the things I need to focus on? " So these AI agents need to interact with the underlying IT infrastructure layer, which basically is what ManageEngine is. So we have a huge portfolio of products, starting with a unified service management.

So like Help Desk has all the information about every single event that's happening in your organization, be it users complaining or talking about something that is not working or your event management system kind of firing incidents, saying whatever application or device went down. So if you actually think about it, your help desk becomes a treasure trove of information that your AI agents will need in order to answer the questions that you ask in terms of what do I need to pay attention to. So now what AI agents are really helping you do is save a lot of time because you don't have to go in and figure out what happened in the last 24 hours.

AI agents, especially AI, is very powerful in terms of consuming a lot of data and summarizing things for you in a very short span of time. And that is an area a lot of our customers are also exploring. But for these AI agents to give you reliable information that you can trust, you need that trusted infrastructure layer which ManageEngine is known for and will continue to refine.

Now IT folks spend a lot of time looking at dashboards. Right. In the age of agentic AI- Mm-hmm ...

is that still going to be a static dashboard that I look at or is it going to be more like I ask a question and then the system kind of dynamically generates the dashboard at the moment that I need to solve this particular issue and then goes away? Right. So I'll answer it this way.

Now I firmly believe dashboards are not going away. The reason why I say that, but again, the relevance of dashboards are shifting. So for example, if I'm running my IT operations, there are certain KPIs that I always want to have an eye on.

There, it doesn't make sense for you to send in an army of AI agents to generate those KPIs every single time you want to look at it. So your dashboards are already a curated list of KPIs that you care the most about and you can just look at it. It's always there without you triggering a workflow to generate that each time.

So that is where I see dashboards fitting in in the future, and I would say even to an extent even now. But where AI agents will come in is, let's say you want to get very specific. So you're looking at some KPI, it looks a little off than what it usually is.

You want to know why it is the way it is. So in those situations you can use AI or most people are using AI to kind of figure out or to answer the question why it's looking the way it is. So I would say dashboard would be your layer where you have your important KPIs whereas the analysis and the inference is being offloaded to AI a lot more because like I said, it can crunch a lot of that information, pick up the important pieces that you need to pay attention to.

It also seems like there's going to be different classes of AI agents with different levels of autonomy. Correct. I think as I understand this, I might have an AI agent that I'm asking to go do something, but there's also going to be AI agents that are kind of running semi-autonomously, continuously monitoring something, and then when something happens they'll send me some sort of alert.

Yeah. Is that a fair assessment? Yeah, it is.

So you'll always, as humans working on an IT infrastructure, we put on different hats when we are working on IT without even actually thinking about it. Sometimes we are the guys who are looking at data to ensure everything is right, just doing a monitoring role. Whereas sometimes we spot something that's wrong and we go in and want to make some changes to it.

So then you are someone who's acting on something. So what I see agents becoming now are extensions of those roles that we are performing already. So you'll always have agents which are doing a bunch of different things.

Certain agents will only be doing something like a summarization, whereas some agents would be doing semi-autonomous work where it goes in, finds something that you need to pay attention to, not just tells you, "Hey, this is wrong," but also comes up with some kind of suggestion saying, "This is what I'm observing. " And you have that recommendation. So then you go in and decide to act, whether if you need to follow the advice that the agent is giving you or look at it in a different direction.

And then you have fully autonomous agents. But in my experience of talking to our customers, there isn't enough trust at the moment to let it run completely autonomously. We might get to that stage eventually, but I see organizations being okay with doing completely autonomous operations as long as those operations don't have a huge impact on your IT infrastructure.

So yeah. So how will this impact the IT teams themselves? Because I think it's no secret that there's a lot of fatigue, there's a lot of monotonous stuff that goes on all the time and some people say the thing about IT is it's like being in the army.

It's very boring and punctuated by a couple of moments of sheer terror. Yeah. How does that job function evolve and will it arguably become more fun to be in IT?

Yeah. I think in a way it will. So every time I speak to a CIO or a CTO or an IT manager, they have two different angles that they have to look at when it comes to IT.

One is ensuring everything is up and running. The other aspect is doing something more strategic where they're evolving from what they're already doing to even something like, let's say, moving to a more proactive detection of issues. Now what most IT teams are doing right now is something breaks, they have to go fix it.

Something breaks, they have to go fix it and it's a very monotonous cycle. That's why IT might be very boring. Now if you're able to automate some of that stuff like Even let's say you have an issue with your laptop.

Now you go to your help desk portal, and when you're typing in your request saying, "You know what? " If the help desk software can detect, okay, this is the kind of issue this person is having, I'm going to suggest a solution. So you look at the solution maybe before you even log a request, let's say you try those solutions that is being suggested by the help desk platform, then you could have a situation where you're not even raising that request because AI basically helped you resolve that request even before you log it in.

So that will free up a lot of time for the IT teams moving forward. So they won't be doing the same old, "Oh, I'm going to fix your printer. Oh, the printer doesn't have a tone," or that kind of job.

That will become more automated, wherein now the IT team can focus on more strategic things. So what can you share about the product roadmap for ManageEngine? Because I'm sure a lot of folks are at that moment where everybody kind of knows that things are changing, but they're not quite sure how.

Yeah. So the approach that we're taking with the ManageEngine portfolio is that we want to put in, of course, there's no software story today without an AI story, right? But what we are doing is we are picking very specific areas where we should put in AI, which will actually deliver value to our customers.

So the example that I was just talking to you about where when you're going in and logging a request, the system is reading what your trouble is, and proactively suggesting a few things that you can try. So that helps by reducing the number of tickets that are getting raised in the first place. And if you look at our observability portfolio, root cause analysis is a big issue.

There's 100 different events firing, and then for you to go figure out, you know what? Some configuration and some router change, which is causing all of these events, takes a lot of time. So we're putting in AI to look at all those events and figure out what is the actual root cause of the problem.

" It knows your infrastructure. It understands every single data point coming in from the various ManageEngine products, and it's able to give you a list of things that you should probably look at even before something breaks. So we are, again, a very value-driven organization, in the sense what we try to give our customers is value for the money that they're paying.

So that means we also have to ensure that we build something that is of value. So we're not throwing AI at everything. We're picking the right use cases.

A lot of the time, what we see with a lot of our customers is they just want to implement AI without knowing what is that one particular problem or use case that they're trying to solve. So we are taking the approach of, okay, let's look at 10 or 50 different things our customers are doing with our products. Let's see where it will actually add a lot of value if I put AI in there.

So that's kind of our defining approach, or the approach that defines how ManageEngine is looking at implementing AI in our software. All right. I think it's not much of a secret that most people probably don't use 80% of the functionality that exists inside of a given product.

Yeah. Is that going to change in the age of AI? Because it seems like the AI agent will invoke a lot more capabilities than humans will, just because the humans are limited by a certain amount of cognitive thought processes, and the AI agent has all the time in the world to do anything and explore everything.

So is the nature in the way that the software gets invoked, the utilization going to just increase? Yeah. What you said is absolutely accurate.

We do a lot of customer visits. We sit down with the people that are using our products day in, day out. " And they describe something they want to do, and you say, "You know what?

" So what these agents know, so we at ManageEngine, our AI ecosystem is under the ZIA umbrella. That's Z-I-A. So what it can now do is, if you describe the problem that you're facing, because it has knowledge of the entire product and how it works and the different functionality, it will be able to suggest, "There is a functionality in the application that will allow you to do what you want to.

" So, yes, to answer your question, that will definitely help organizations utilize the product a lot more because now they don't have to sit and understand every single piece of how the whole thing works because there is a lot of functionality. And most IT teams are very, very lean. They have probably three or four people, and like I said, they're buried with mountains of issues that they're dealing in on a day-to-day basis.

They never have the time to go in and see what is actually in the product. So I think, yes, the whole advent of AI would help organizations get the most out of their products. Yeah.

What's your best advice to the folks that already have your software installed about how to get ready, or for those who don't, for that matter, but how should they be thinking about getting ready for this? Because the way we manage IT is fundamentally changing, but I'm not quite clear to your earlier point, just throwing AI at something is the right approach. Yeah.

So, my recommended approach, and this is something I'm not saying just for the interview, this is something that I say to the client that I meet always is, start with three problems that are fundamentally slowing you down in everyday IT. Start with those use cases. Are you having trouble with too many requests?

Then that's a problem. Are you at a point where you're not being proactive, you're always reacting to something that has already happened? Then that is a problem.

So look at what you can do that will fix that, because that will free up a lot of time. Even if you address the top three issues that you're facing with IT, and diligently put AI in there to solve those problems, then that by itself is the biggest benefit. So I think in summary, don't ask what AI can do.

Start with asking what you want AI to fix first. And that's a good starting point that I always recommend to our customers. Awesome.

And Rakesh, thanks for coming on. Thank you so much. It was a pleasure.

And thank you all for watching that episode. And I think the one thing that's clear is, hey, the way we manage IT will never be the same again. We'll be back in a minute.