Transcript:
Pete Wright:
Today we have the privilege of speaking with Murali Balcha, the visionary founder and CTO of Trilio. With decades of experience in enterprise software, Murali has been at the forefront of innovation and data protection for cloud infrastructures. Today we explore this journey from founding Trilio to securing major industry clients and navigating the dynamic world of startups. I’m Pete Wright. Welcome to Trilio Insights. Murali Balcha, welcome. I’m so glad to be able to sit down with you today.
Murali Balcha:
Thanks a lot, Pete. Thanks for having me here. I’m very excited.
Pete Wright:
I’d like to go back in time, if we would. As folks know from the title and show notes of this very podcast, we’re revisiting a post you wrote on the Trilio blog about some of the founding days of the company and enthusiasm about securing funding and watching the company grow. I want to go back to your time at EMC. You had been working there for a while and you realized at some point that you had an idea that you felt was startup-worthy that you can contribute to the industry. Do you remember that time? What was that like?
Murali Balcha:
Yeah, I still remember that vividly because it’s hard to forget, and no matter how much time has passed. So it’s almost 10 years since we started. It’s 11 years, wow, 11 years since I started the company. It’s long journey and twist and turns along the way. Any startup company, the founders thought they have phenomenal idea that it’s going to change the market and they thought it’s going to be straightforward to basically get to success. But once we get in there, then it realizes it’s not that straightforward. But I think every founder, they’re driven by gut feeling. It’s not market research. It’s a gut feeling that okay, they have a better idea than what the market has right now. And that drives any founder to basically leave a cushy job and go down this path of starting a company, becoming an entrepreneur. And it’s the same thing with me.
So I’ve been with EMC for almost 15 years. And EMC was only 6,000 employee at the time, very small compared to what EMC growing to. And so I was part of pretty much the whole evolution of storage area networks and storage and how EMC had led that technology. And then at some point it became one of the four hearts of that internet.com boom. And EMC acquired virtualization platform like VMware. And I was part of that and I led a lot of virtualization initiatives across EMC, across all the product portfolios. And then the cloud is kind of nascent at the time. It’s been on the radar of EMC and there is a lot of excitement about the cloud and a lot of companies are seriously considering moving to cloud.
Pete Wright:
A lot of uncertainty too, right? Do we, don’t we, do we, don’t we? On-prem, cloud. Big transformation.
Murali Balcha:
For any technology, transformative technology, there will be hesitations, skepticism, the privacy concerns, there are a lot of things that will be there, but then they’re all one over the elasticity and the evaluation of the technology. So we thought that’s a good inflection point for us to do something in the cloud. And that’s when I started the company with a couple of my friends, left EMC and started the backup and recovery for the cloud workloads. So that’s where our journey started. And we started focusing on nascent cloud platform, an open source platform called OpenStack at the time. The industry was excited about it, lot of vendors throw their weight behind it and thought it’s going to go places. So we left EMC and started the company solely focused on OpenStack initially. So that’s the genesis of the company. And then we had a lot of surprises along the way.
Pete Wright:
That’s where I would like to go next. How soon did you realize that the company that was in your head, you and your colleagues head when you left EMC, was going to cement into what it is today versus realizing things are not going the way we intended, we’re going to have to pivot?
Murali Balcha:
The use case that we took, which is backup and recovery, that is in industry they call it day two problem. So that means you need to have a production workloads already migrated to a platform. And that’s when people start looking at backup and recovery and resilience and that kind of features that they want to basically protect those workload [inaudible 00:05:51]. So when we left the company and started this and OpenStack was evolving, there are a lot of hype around it, a lot of companies talking about it, but the production workloads are far in between. There are not many production workloads and most of them must take less workloads If they crash and burn, you just restart those workloads. No one really thinking about backup and recorded for this workload. So we are very early when we started, so we did not gain much traction for almost two years. Though we are focusing on developing the product, but we could not really land a paying customer until 2016. That’s when we had the luck of closing a deal with one of the largest telecom providers in the USA.
Pete Wright:
This is where I get to have my mind blown. You enter a space that to me implies we need, in order to even have a proof of concept, we need some ridiculous infrastructure to be able to even test this thing. And also by the way, we’re going to go sell it, at the same time we have to sell it. And I need to know how you close that gap. How do you figure out those problems in order to be able to prove that it’s worthwhile?
Murali Balcha:
Yeah, startups means you had to be very, very scrappy. Right? At least in this age, you have the cloud and then you can quickly put together a reasonable amount of resources in the cloud and pay them on a monthly basis on a use basis.
Pete Wright:
And you’re doing this, you guys, you’re self-funding at this point, right? You’re doing this out of pocket.
Murali Balcha:
Self-funding. And then so what we did was we basically went to eBay and then bought half a dozen servers, decommissioned the dev servers and put them in the basement. That became our development platform to basically build this solution.
Pete Wright:
Outstanding. This is the juice, this is what I needed to hear. All right.
Murali Balcha:
So the servers were very cheap, $200 a piece. And we bought them and put them in the basement and connected to the network and that’s where we started building our product.
Pete Wright:
But you were also though, leaving EMC, you were running organization-wide initiatives, technical initiatives, right? You went from running very large projects to, okay, now we’re going to start at zero again and we’re going to take it out there. On a personal level, at what point did you realize, I think I’m ready to do this from scratch and take on that responsibility?
Murali Balcha:
Yeah, well, when you leave, you always leave with optimism that you’ll somehow end up in funding, but you’ll have a rude awakening because raising money is not easy. It’s very, very frustrating because you need to find investor who basically believes in the same mission that you have.
Pete Wright:
And it’s also wholly different from the work you’re trying to do day to day.
Murali Balcha:
Yeah, right. Absolutely. You need to wear so many hats. Nighttime, you need to be a programmer tester and everything. And then at daytime you need to basically put together a presentation and then try to find investors who can listen to your pitch. And it’s not an easy thing. Especially for the first-time entrepreneur who doesn’t have any name recognition in the industry, not many contacts in the VC community. It’s extremely hard to do it. Fortunately we bumped into David Safai and then he was a VC before and that’s when it started clicking really. Up until then we’re like a bunch of technical folks just dreaming things but don’t know how to make the business out of it and how to get the funding, how to build a company.
Pete Wright:
You get that sale, that first customer in 2016. What does that do to the business?
Murali Balcha:
Most of the startup companies, they’re lucky to basically land small small customers. Even 10,000 and 15,000. Kind of those deals are very hard to come by. But I think this telecommunication company was building open OpenStack cloud internally, and there are nothing in the market that can provide the kind of solution that they’re looking for that is kind of neatly built into the cloud and where each individuals can [inaudible 00:11:03] the backup and recovery out their workloads. And none exists, right? So they really liked the approach that we are taking. They were willing to basically partner with us and willing to basically go take the journey along with them. That was a huge deal. It was a seven-figure deal, which is very rare for a startup company to land that kind of deal.
Pete Wright:
Yeah.
Murali Balcha:
So that basically helped us set the company in the right direction and helped us raise the out series.
Pete Wright:
Yeah, runway, is that the right word? I believe it gives you runway. Right. Okay. I have talked to the aforementioned David Safai, very charming. I have talked to Rodolfo, I’ve talked to Kevin. We have talked extensively on the show about what Trilio is. I am eager to hear from you what is it that makes Trilio the company and Trilio’s solution unique?
Murali Balcha:
Backup and recovery is something that it’s not sexy, but every company wants it and especially backup and recovery was meant for some data corruption or some hardware failure. So those are the use, to protectiveness those, something that naturally occurring that usually people use it. But then the use case broadened, especially with the cyber attacks and the ransomware attacks and also multi-cloud scenarios. So there are scenarios that is driving the backup and recovery market. The backup and recovery, they were there like 30 years back. There are backup and recovery products that pretty much does backup and recovery individual files and then backups to a tape and then recovery. So those are all not really very dynamic in nature, but when you look at how the workloads are created now, they are highly scalable and then your solution has to grow with those applications and how the applications are evolving.
In back even 10, 15 years back, when you deploy one application like Oracle, right, it is something that you deploy with 10 terabyte data and then it is stationary, that application stays there for years without much modifications. Most of the backup solutions, they are addressing that market, the applications that are very static, you only back up the data and then recovery is recording the data back to that application. But when you look at the application, the dynamic nature of the application, how they horizontally scale in terms of memory storage and everything, that doesn’t really make sense. So what you need is a much nimble approach where the backup and recovery knows how the applications are deployed at that moment in time and then take the entire snapshot of the application and then ability to restore not only the data but the application to the point in time where it was taken back up. But not only that, these applications can be taken in one cloud and they can be restored on a different cloud.
So this systems should be able to help you do the restore in different environments. Right? So this whole dynamic nature that brought in, that needs a completely new approach to this problem. And that’s where we kind of shine in that. And ease of use. For example, for VMware side, you have a central administrator who is responsible for backup and recovery of all the applications. The end user probably doesn’t know about it because end user is application developer. We only focus on making sure that application up and running. But in case the application goes down, they call the backup and recovery administrator to restore the backup.
But in a cloud environment that is not positive. The cloud administrator is setting up the entire cloud for you like AWS, right? And the burden is on you to not only deploy the application to make sure that application you are protecting those applications. So this functionality that as a dedicated central administrator now move to end user now and that is end user responsibility, how to protect your applications and record those applications. This new evolving paradigm, we address that very well compared to any other vendor in the market.
Pete Wright:
Yeah, I want to go back to sales because everything you’re describing to me, I remember the first time I was introduced to the conceit and I had another one of those mind-blowing moments because I couldn’t conceive of how that was possible and let alone how that was easy. What are the reactions that you get from folks who are just introduced to Trilio for the first time who have a direct need?
Murali Balcha:
There is a competition obviously. Right? And when you look at, they will have backups. They don’t take what we basically put it on our website. They have to verify firsthand that our solution is what they wanted. And we are, the backups usually involve the existing players and then us. Our name is not very well known in the industry. There are a few people who are in the open stack market, in the cloud native market and the right-hand market, they know our name, but it’s not well known outside. It’s not mainstream. But then they basically realize doing their own research that we have a differentiating offer and then they do a backup with that. And in almost all the cases we win because of the differentiation that we have. The response usually is pretty good because they go compare even feature by feature and then they choose Trilio.
Pete Wright:
You’re modest. I can hear you trying to stay modest, but that differentiation is extraordinary and really fun to see at work.
Murali Balcha:
Yeah. And some of the backups, they usually do the backups with three or four vendors and they run for six to eight months and then they pick us. And that means a lot. These are big, big companies. They have every vendors at their disposal to basically evaluate each and every offering and then still choose Trilio. That means a lot.
Pete Wright:
We’ll speed up the timeline. I want to bring us to today because I am interested in what is next from your perspective as CTO founder, one of the leaders of the organization. As you look ahead on this sort of six month to five year timeline, what does the space look like, the technology space look like for Trilio? Where are you, obviously without giving away too many secrets, where are you investing?
Murali Balcha:
There are a couple of places we’ll be investing. One is there are so much changes that are happening in the market. One thing that is changing is Broadcom acquiring VMware. That is causing a lot of concern in the industry and a lot of customers are looking at some alternatives to VMware. The other thing is that there is a technology inflection point in the industry. Virtualization is going to remain for the foreseeable future. Most of the companies who have critical workloads are still running on virtual machines. And the work [inaudible 00:18:47] behind it is the VMware obviously.
But then VMware is a closer proprietary software that is very robust, rock solid, gold standard. But then if we look at the agility that people are looking for with the DevOps, with a continuous integration with the multi-cloud and then the vibrant ecosystem policy-based main. So there are a lot of things that VMware may offer, but they don’t really, that whole workflow doesn’t flow. And then the community, especially the Kubernetes, the community is very vibrant and they come up with very, very innovative solutions. We see a lot of movement there and we are building pretty good healthy pipeline on this virtualization for Kubernetes and that people migrating from VMware to Kubernetes. So I think that is one area that we want to lead. Right now I think we have a very good solution, differentiating solution, but we want to make sure that we solidify our leadership there. And I think the feature, you cannot talk the feature unless you talk about artificial intelligence.
Pete Wright:
I was wondering when that was going to drop.
Murali Balcha:
Yeah. So Gen AI, generative AI like ChatGPT, I think that we strongly feel that’s going to transform how everything is done in the humanity. So far, how we have been interacting with the systems and everything changes because of the generative AI and that’s going to affect how people manage their data centers. And that will affect it in a way that we haven’t seen up until now. So this is going to touch every aspect of how data centers are managed. And backup and recovery is part of the data center and it will be impacted by that. So we are taking a lot of initiatives to make sure that Trilio products are enabled for the Gen AI use cases. We are very much excited about it. And I think in the next one or two years, you’ll see lot of good solutions coming out of Trilio for Gen AI use cases.
We definitely want to start small and baby steps and then basically evolve into much richer offering. That’s what we are. But think about simple use cases. For example, we have a very, very good UI that can manage the backup and record for multiple clusters. User need to have everything in their mind before open the UI and then go and click through each of these resources that they want to protect and what kind of policies that they can set. The moment you open the UI, it can basically discard the inventory, discard the best practices to protect everything that is there in your environment, and then offer all the possible solutions for you to basically automatically protect this. If they run into some issue and they have an error message and insert your backup, it can basically click and then potentially give you all the possible solutions for you to resolve that issue without calling back into the Trilio for support like level one level two support.
Pete Wright:
Sure.
Murali Balcha:
Those are some of the things that can be easily done. Right?
Pete Wright:
Yeah.
Murali Balcha:
And the other thing is a monitoring and alerting. I think the business analysts or anyone, when you look at the monitoring, the dashboard that people provide, they have predefined KPIs that you have in the monitoring. Right? You can monitor number of VMs or number of CPUs. So these are all predefined. Either you like those metrics or if you want to extend beyond that metrics, then you have to call back the vendor and say, “Hey, can you add more metrics, more metrics?” With the Gen AI, you don’t have to do those things. You have a very, very simple one-line dialogue box that says, give me this information that I’m looking for, right?
Pete Wright:
Right.
Murali Balcha:
Tell me all the backups that have failed in the last 20 days, right?
Pete Wright:
Yeah.
Murali Balcha:
And tell me how much stories I’ve been using for my backup. Tell me based on this usage, when do I need to refresh my backup target? When do I add more [inaudible 00:23:39] to it? So this natural way of communicating to your system and then get the feedback, that’s going to be very important. So there is nothing pre-determined at this point. It’s all about the user is free to ask whatever they want to the system and then they will get a meaningful response to you.
Pete Wright:
Well, I think you said the most important word, getting a meaningful response to you. I think right now, when you look at the level of end users managing these systems, to what degree do they trust AI to help them do the work of their work versus do they experience, how long will it take us to be beyond the era of hallucination in these sort of micro functions? And trust is a big part of it.
Murali Balcha:
Trust is a big part of it, and I think that’s where we need to focus and we need to have a little bit of guardrail saying we can’t fully expect AI, AI is going to solve, but at least provide more helpful hints to the user on how they can manage the systems. When you look at the AI, it is not coding anymore. Coding is so rudimentary these days. You don’t need to code, but it’s all about prompt engineering, what they call it. You express your intention with a natural language as precisely as possible so that, yeah, I can generate the right responses to you. So going forward, the software, when you ship a software, you probably don’t ship software. You basically ship the new set of prompts that are tested.
Pete Wright:
Right. Right.
Murali Balcha:
So it is a text file that you are shipping in most cases, if you basically do this, do this right. It’s not the code anymore, right?
Pete Wright:
Yeah.
Murali Balcha:
So it’s going to transform the industry.
Pete Wright:
Not long ago somebody discovered the text file that is the prompts for the new Siri that’s coming out. Things like be accurate, really focus on being accurate, things that give us a good chuckle. Those of us who use it day to day. I’m curious, what is your personal use case for AI right now? Just sitting at your desk, getting things done, how do you use it? What’s your platform of choice?
Murali Balcha:
Yeah, I use ChatGPT a lot, and I use Meta AI a lot.
Pete Wright:
Okay.
Murali Balcha:
In most cases. I could do Google search and spend half a day to figure out what exactly that I’m looking for, but if your prompt is precise enough and ask the ChatGPT or Meta AI, it really gives you a pretty good response, right? It’s not only describing simple domain specific question, but also sometimes you can ask, “Hey, can you back it up with some [inaudible 00:26:47].”
Pete Wright:
Sure
Murali Balcha:
And how you can demonstrate that. And they’re pretty accurate.
Pete Wright:
Yeah, I’ve had the same experience and we play with them all around here, Perplexity and Gemini, and they all seem, I think we may be entering an era where that lack of trust from a lot of folks is getting in the way of, will be a hurdle for getting things done now that we’re sort of seeing a lot of the early issues resolve themselves. And it’s moving very, very fast.
Murali Balcha:
Very, very fast. And then the amount of research and the amount of progress that we are making in AI is mind-boggling. You might have heard what’s called agentic workflows, AI agentic workflows, and Salesforce threw big weight behind it, and other companies are also going in that direction. When you look at big problem, right? So let’s think about an organization. You want to set up an organization, you want to have 10 people, one looking at the marketing, one looking at the legal, and one looking at engineering and one looking at something else.
Now each one has a role. So that means when you do the job description, you precisely describing each and everyone’s role is. Now with the agentic workflow, you are basically defining these roles to one each agent. Right? Now you have a marketing expert agent, that virtual assistant that you… And then you define the workflow and how these agents communicate with it and delegate each other and get the final task completed. One other area is we want to contribute ourselves and so that we can have our agents published. So when someone is building a more unified workflow, AI workflow for their organizations, we are just part of it.
Pete Wright:
You have a Trilio agentic workflow?
Murali Balcha:
Yes. So those are some of the things we have [inaudible 00:29:07].
Pete Wright:
Outstanding. Last question and then I’ll let you get back to your day. As a founder and CTO, Trilio continues to grow around you, how do you see your role evolving with the company over that next same period, say five years?
Murali Balcha:
I think that’s a very good question. I want to be a change leader. Right? I have a very good team from engineering point of view, everywhere. I’ve been able to guide the Trilio so far in the last 10 years going from OpenStack, going from [inaudible 00:29:43] to Kubernetes. And then now with the Gen AI. I still feel like there is a lot of room for me to grow and bring more differentiating offerings to the market because not just dreaming up new technology, but also dream technology and the products that are impactful, and then revenue generating for the company and position the company in a leadership position. So that’s always on top of my mind. How do we make this company bigger? How do we basically change the offerings so that we have bigger market share to address bigger problems to address? So hopefully my efforts will be fruitful and company will grow much bigger than what [inaudible 00:30:47], right?
Pete Wright:
I can’t tell you how grateful I am for your time today, Murali. Thank you so much for sitting down with me and sharing the story.
Murali Balcha:
Yeah, thanks a lot, Pete. I greatly appreciate your time.
Pete Wright:
Thank you everybody for downloading and listening to this show. As always, we appreciate your time and your attention. We encourage you to learn more. Just swipe up in the show notes for all the links that I will put in there associated with this episode with Murali. If you have questions, there will be a link in there for you to ask them. As ever, on behalf of Murali Balcha, I’m Pete Wright and I’ll catch up with you next time right here on Trilio Insights.
