3173: The Future of IT Operations – Introducing Fabrix.ai and the Agentic AI Era

[00:00:04] What happens when a company decides to redefine its identity? That's exactly what I witnessed live at the IT Press Store in Silicon Valley. I was in the offices of Cloud Fabrics and they quickly revealed their transformation into fabric AI. But rather than just hearing about it later, I had the opportunity to capture that moment as it unfolded.

[00:00:28] So what you're going to hear today is the inside story of a company that is not just reshaping its own future, but also tackling one of the biggest challenges in IT operations. And what I'm talking about is bringing together observability, AI-driven automation and data intelligence. And doing that in a way that makes enterprises more autonomous. And at the heart of this transformation is today's guest.

[00:00:55] He's a serial entrepreneur and IIT alumnus whose journey has been defined by a passion for solving complex technology problems. He's built and exited multiple startups. And now with Fabric AI, he's setting the stage for the next evolution in enterprise automation. So today we're going to explore his vision, the problem that he witnessed firsthand in IT operations

[00:01:21] that initially led to the birth of Cloud Fabrics and how he's now executing that vision through innovation, strategic partnerships with some pretty big industry giants like Cisco, Splunk and IBM. And also learn more about his leadership style that has fostered a loyal team across multiple ventures. But beyond the boardroom, who is today's guest? What drives him outside of work and what lessons has he learned from his journey?

[00:01:49] Let's get the full story behind the technology and the person leading this transformation. That's right. Time to get today's guest on. So a massive warm welcome to the show. Can you tell everyone listening a little about who you are and what you do? Thanks for having me. I'm Raju Datla. I'm co-founder and CEO for a company called Cloud Fabrics based out of Pleasanton, California. Now you've had a remarkable journey from your early days at IIT to becoming a serial entrepreneur.

[00:02:18] I was Googling you and your team. Saw there's a bit of a connection to Cisco about 10 years ago as well. So what experiences shaped your approach to building and scaling successful companies? Because you've been in the industry sometime now. I had an opportunity to work with one of the best companies in Silicon Valley, including Cisco and several great companies in India. Working with the Silicon Valley mindset and the companies where you have an opportunity to take the risk

[00:02:45] and you work with the great people in the valley. That kind of helped us early on to take challenges. If you see a problem, don't wait for somebody to solve it. Go solve it. That kind of mindset helped us to, you know, do the things that we have been doing it for the last 20 years. Now, Cloud Fabrics was founded to address the increasing complexity of digital businesses.

[00:03:10] So I'm curious, what problems did you initially see in IT ops that traditional data management frameworks couldn't solve? And how has that evolved to present day and indeed your rebrand of Fabrics AI? When we started the company Cloud Fabrics, one of the problems at that time, you know, almost 10 years ago, was that all the tools, monitoring tools, now what we call observability tools or silo. You know, we have storage monitoring tools, network monitoring tools, compute monitoring tools,

[00:03:39] and then APM or application monitoring tools. They were only looking at one of the layers in the stack. What we wanted to do was the full stack observability, understand the entire stack in a traditional application space, all the way to the physical compute and virtual infrastructure, understand the relationship between the stacks, technologies, compute storage network or cloud, hybrid cloud, how they are interconnected, understand the context and dependency,

[00:04:06] and then build a more intelligent visibility solution, a solution that understands the dependency and analyzes the problem in the context of the stack rather than in isolation. That's what we started with basically when we started. How it evolved today is now observability is more pervasive, and also you see that applications changed from two-tier, three-tier applications to microservice applications today.

[00:04:31] A lot more dynamic, a lot more challenging to get the real-time visibility of that one. So we saw this transition from early cloud to more cloud-native applications, and still we have a lot of applications still running in data center and on-premises applications. So you're not motivated to the cloud, it's still there in data center. So the observability challenge got a lot more challenging. You still had to monitor the old applications, the old tools, as well as the new generation one.

[00:04:59] And you had to do it in a cohesive fashion. That's what we have been doing at Cloud Fabric. We caught up with the transition industry to do more cloud-native as well as legacy applications. And here in 2025, we're entering another chapter now. Towards the end of last year, Agentikai, everyone was talking about that. It felt like there was a big change coming. And your concept of the autonomous enterprise is built on three pillars.

[00:05:26] Data fabric, AI fabric, and automate fabric. Can you break down what each of these are essential to the problems you're focusing on solving with this new rebrand of Fabrics AI now? So with the Agentikai, what you're doing is you're using the AI to solve the problems, not just being an assistant, not a co-pilot, where you're asking a question and giving you some information. It's up to you how to consume the response you got from AI. It's pretty good in most cases.

[00:05:56] We have been using it the last two, three years with chart GPT, perplexity, and a lot of other ways. But this year and last year, we have been trying to go to the next step. How do we let AI do the work wherever it's possible in an automated way? For that to happen, just having an AI model is not enough. You need to have access to the real-time data. You need to have access to the automation.

[00:06:21] AI can run the automation to do certain things that a normal human would do or human would run. Without that automation and data, the AI can only look at what it knows from the trained model. Or you can say you can run it, but it doesn't know how to run it because it's just a disconnected computer. With the data fabric, we are able to get the data from hundreds of tools.

[00:06:45] Typically, our customers have anywhere from 50 to 100 data sources, legacy to more traditional and cloud native. And they're all very dynamic. This is not something that you cannot talk to 50. AI cannot talk to 50 different tools. We solve that problem, bring all the data, distill the data and put it in a fabric so that AI can access that one wherever it needs to.

[00:07:10] And then we also brought in automation into our fabric where you can connect to different tools and run certain things, verify certain things. Is the problem still the problem exists? Is the data still up to date or is it different? Is the problem solved by some other tool already before I'm trying to solve? Now, so many things a human would do, now automation has to do. Now, AI is the mind that's sitting and saying, I need this data. I want to verify this one. I want to run this automation.

[00:07:39] Cut back with the data and say, is this fiction now? These are changed. AI is able to orchestrate the data and automation. Now, the agents created with these three things together are the powerful agents that can actually solve the problems. Versus pure generative AI, it is only telling you what it knows from the train. It also eliminates the need to keep training the model with the data.

[00:08:05] I think we are moving away from the concept that we have to train the LLM models with the data. You're only going to train with some information how to pull the data in real time. Mark Nussle actually trained the data using RAAG models. So, we brought this together in a way that now agents can be developed and deployed. They work autonomously because we have all the necessary pieces together. Intelligence, data, and automation, ability to execute and run tasks.

[00:08:34] That's how we are adapting to this one and providing agents to our customers. And as I said, it really feels like another chapter for you guys because Cloud Fabrics, over 10 years, has gained recognition as a leader in AI ops, observability, and automation. So, what key product innovations have been so critical in executing this new vision? A new vision that's going to set Fabrics AI apart.

[00:09:00] So, one of the innovations that we have done about five years ago was robotic data automation fabric. It's a way to connect the data, pull the data from different sources, transform it, enrich it with the network topologies or application topologies, our customer context, business context, and then use it for AI ops, AI-driven IT operations and correlations.

[00:09:23] Now, for the agent AI, one of the key evolution that we brought it in is taking the LLM and putting the necessary guardrails around that one. Making it more predictable, not hallucinating based on how the question was asked or something else that affected that generation process. How do you make it repeatable, predictable for a given condition? How do you put guardrails around it?

[00:09:50] How do you put boundaries so that it doesn't do something unwanted? That is always the fear of humans against AI is will it have its own mind to do something unwanted? Even though I want it to do certain tasks, goes and bear something destructive. We have to put the guardrails around it. We have to have the humans in control of that one, of the AI, not the other way around. So, we have developed the framework platform that ensures that AI, there are guardrails around it.

[00:10:17] It's predictable, not randomly doing certain things. And your user has the control of whether AI will do it or not do certain things. And we also brought in a very intelligent way to bring the automation and give the power to AI. You can run these things. You can verify this. Not the, we are telling you to do it. You decide AI that I need to verify this before I do this. So, we put enough guardrails so that it doesn't do any negative things.

[00:10:45] But we are letting AI do automation and data gathering and making a decision and then perform the decision. Throughout the whole process, humans have the control. Our customers, you know, for the lack of term, their agent masters, they are the one controlling these agents. And saying, okay, you can go live now. I'll keep monitoring. So, that is kind of the innovation, the transformation that we're bringing together data, automation, and, you know, AI with the guardrails.

[00:11:15] Human oversight. What we believe is the game changer to get to the agentic AI. And we're recording this today in your offices. And as I was watching the demo earlier, you were talking about automation and how it would automatically raise support tickets, etc. And I thought, this is great. But as an ex-IT guy, what about alert fatigue? And that was always the biggest deal there. And I thought, well, how's this going to deal with that? And then you quickly dropped a stat in the next slide.

[00:11:44] It can reduce it by up to 95%. Can you tell me a little about that? Absolutely. Absolutely. So, with the, you know, evaluation of a lot of monitoring tools, a lot of microservices applications, we have more observability data. Which means we are generating a lot more alerts. You know, almost millions of alerts per day in most of complex environment. What that means is, that's a human cannot look at those alerts and react the way in the past we used to react. Oh, server is down. Let me go figure out what happened.

[00:12:13] Now you are getting thousands and thousands per minute or hour that humans cannot react. Either you are ignoring them or you are just doing some brute force method. You know, I get 10 alerts on this one, then I will react to that one. What's happening is you are wasting time on alerts that are not real or not important or missing the alerts that are actually indicating a big problem. How we have solved the problem is looking at the application dependency mapping full context of the entire application.

[00:12:40] Correlating everything, correlating everything, correlating everything, all the alerts coming in and understanding how these alerts impact a business critical service or they don't impact a business critical service. And doing the D2, you know, all the anomaly detection in the alert volume processing, we are able to bring down that to 95% reduction of the alert noise.

[00:12:59] And then even then, when we create an incident or a ticket, we are still continuously monitoring what's going on with alerts and either update the incidents or notify the user or create a new incident based on the user's preference. So we are following through the lifecycle of the entire incident to make sure that alerts noise doesn't get to the users. Because they're only dealing with the real incidents without missing the critical incidents.

[00:13:25] And I always think you can tell a lot by a tech company, by the kind of partnerships and collaborations I've got. And when I was doing a little research on yourselves, you've built strong partnerships with major players from Cisco and Splunk to IBM. I'm curious, how have those collaborations shaped your growth and indeed the evolution to Fabrics AI? And what do they mean for the future of AI-driven automation, do you think?

[00:13:50] You know, when working with the large telcos or enterprises, you know, it's as a startup and going and trying to sell the next generation journey and autonomous enterprise. It's hard. What we have taken the approach is, we have taken a consultative approach, working with the partners that already have the mindshare of these customers. Like Cisco or IBM or even Splunk or a lot of other partners that we work with across the world. They already have the ear and mindshare of the customer.

[00:14:19] They already know the customers from all the pain points. When they look at it as CloudFabrics or eFabrics.ai now, we can solve these problems that they connect and say, we want to bring in Fabrics.ai. We can help you solve this problem. We are in the discussion together. That's what Cisco has been doing with a lot of global telcos and very successful. You know, both Cisco and us have been very successful in delivering to global enterprise and customers.

[00:14:46] And customers are extremely happy with this partnership. Same thing we have been doing with IBM and other customers. So, go to market through large SIs and large partners. Helping us, the customer, the partners, and us. I think we have accomplished with a lot less, you know, with the help of these partners. Otherwise, we would have had to have a lot more of investment and so many other things. And your technology is incredibly impressive.

[00:15:15] But as I said at the very beginning of our conversation, I also noticed when I was Googling your team and where everyone's came from, was that how they've been around for a long time at different companies like Cisco, for example. So, from a business leadership point of view, you've built a team that has followed you across multiple startups. What's your approach to tech leadership and company culture that fosters such strong loyalty? Because there'll be a lot of business leaders wanting to learn how you've mastered this, I would imagine. Yeah, absolutely.

[00:15:45] We have the fortune of working with our team for a long time. Some of us have been working for 20, 30 years and almost 30 years, you know, in different companies. So, we work together very well. That's because we are a very collaborative culture. You know, we don't have a typical hierarchy that, you know, UTM companies do and typical management culture where, you know, managers are telling, the workers are doing, engineers are doing.

[00:16:15] Here, everybody does the work by example, lead by example. And, you know, we don't, we all companies want to feel the pain and pressure that comes from our customers and partners, not internally created through our culture. We don't want our employees to be afraid of their managers. We want to be partners with everybody. So, in general, we have hired the leaders that are very good, hands-on and very collaborative.

[00:16:42] And that helped us kind of foster that culture in our, you know, across the board. Now, we are close to 120 employees and we still behave like a 10-employee company because we are very open, you know, have a very open borders. We don't have any offices that you see. We sit across from each other, have very open discussions. That helped us to stick together, enjoy working together. And when we go and start a new company, we all gravitate to, you know, together naturally. I love that.

[00:17:12] And having someone that has founded multiple startups with successful exits, for founders listening, what lessons have you learned along the way? Any advice you'd give to entrepreneurs listening to our conversation today? Maybe they're looking to follow in your footsteps, build sustainable, high-impact businesses. Any advice that you would pass on? One advice I always give anybody that asks my advice about starting companies is, do make sure that you do it because you're passionate about it. You know, not because you want to make money out of it.

[00:17:42] You know, that would come naturally if that is successful. Because one, you're passionate about it. And build the company, you know, in a right way, with the right culture, not take shortcuts. And you're not building the company to sell it, to make money. You're building the company because it's the right thing to do. If it so happens that somebody wants to acquire it, you want to sell it, let that happen. But don't build the company to sell it, but build the company because you want to build the next unicorn, next, you know, big business.

[00:18:12] And we feel very fortunate because we have great partners. At the same time, we are at the doorstep of the great revolution and revolution with Agentiki AI. Wise words indeed. We've all seen many startups out there that are built just to be acquired. And what about yourself? I mean, outside of leading what is now Fabrics AI, what are some of your passions? What are your interests beyond the tech world that keep you inspired, make you want to jump out of bed in the morning and keep this passion going?

[00:18:39] And we are actually, most of our employees and others are very passionate about sports. You know, we follow the sports throughout the world, you know, very closely, have passionate discussion conversations, go attend the games. You know, I have been season ticket holder for local NFL and many games. We go every year, you know, we attend the Super Bowls, all these things. In general, we are very passionate about sports. And that kind of brings us together and, you know, how things we talk about outside the technology

[00:19:08] and outside the office where we still enjoy the time talking about it, arguing about it. And I think that is one of the main passions that we enjoy. Besides other things that we do, like social activities, but that's the number one passion. Brilliant. I love it. Obviously, we're recording this on the day that you've rebranded to Fabrics AI. There's a lot of excitement, a lot of questions out there. What can we expect from Fabrics AI this year? What's next? What can people expect in 2024?

[00:19:36] So, Fabrics AI is going to, you know, kind of take the agent AI to the doorstep of all the enterprises, especially in the IT space. It is available now where they can transform their business and how they look at the data, how they can automate it, that vision that they have, two-year plan, five-year plan, we can bring that to your doorstep now.

[00:20:02] And a lot of our customers love our platform because how it brings the data and everything makes it very seamless. And now that we are also bringing the storyboards, the storyboards are a way to tell the story in a user-centric approach to present the information. And storyboards and agent AI is going to change the way people look at the IT and, you know, how they manage the complex IT environments, applications.

[00:20:30] We are fortunate to have the customers that are now the largest in each segment in the world. And we want all our customers that are going to be in a mid-level or even enterprise smaller see the power of this technology and use it in the next one year to years. Well, I suspect there's lots more to come. We're going to be staying in touch very closely. It'd be great to get you on later in the year, see how things are evolving. But before talking to you today, I've been talking to a few members of your team. I believe you're going to be on the road.

[00:20:59] We're recording in Silicon Valley. You're at several events across Europe. For anybody listening, where can they find you on the road? What kind of events are you going? And also, where should they find out more information if they just want to dig a little bit deeper? Absolutely. Please go to fabrics.ai, our website, and we have all the up-to-date information. But we'll be part of Mobile World Congress that's coming. And before that, Cisco Live Europe and then Cisco Live in San Diego for the U.S., North America.

[00:21:29] And those are the three major events that we are attending. We're going to attend a lot of Splunk user-based, Splunk conference, that we have partner conferences for Splunk. We participate in those. So we work with a lot of our partners, as well as these major events. And look for fabrics.ai, you know, more of these updates on where we are going this year. Well, we could have spoke for another hour on this. There's so much great work that you're doing at the moment with Argentic AI,

[00:21:57] with unifying data silos, and also reducing alert fatigue. I would urge anyone listening to check you out if they're going to any of those events to see you there. But more than anything, thank you for your time today. Thank you, Neil. I appreciate it. Thank you. So as Cloud Fabrics transitions into Fabric AI, I think it's clear that this is so much more than a rebrand. It's actually a strategic shift towards a future where AI, automation, a data-driven intelligence,

[00:22:25] all work seamlessly together to power enterprise IT. And my guest's journey from his early days at IIT to leading multiple successful startups, I think it offers a look at what it takes to build and sustain innovation. And I think his approach to leadership rooted in authenticity, mutual respect, and a win-win mindset.

[00:22:48] I think all those things have been instrumental in not only retaining a dedicated team, but also forging so many strong, huge industry partnerships. So today, the story of Fabric AI is ultimately still being written. But the vision is clear. IT operations need a smarter and more autonomous approach. And this transformation is setting the foundations for what is next. But over to you.

[00:23:15] What do you think about the role of AI will play in reshaping enterprise IT? Are we on the cusp of a truly autonomous enterprise? You know the drill. Let me know the thoughts. Email me now. Techblogwriteratoutlook.com LinkedIn, Instagram, X, just at Neil C. Hughes. Easiest guy in the world to find. So please, send me a quick message. Let me know. But that is it for today's episode. I'll be back again tomorrow with a completely different topic

[00:23:44] and hopefully an inspiring story of how technology is inspiring our life, work and even world. So I will speak with you all tomorrow. Bye for now.