Investors Said "No." Customers Said "Yes." How CoreStack Built an $85M AI Company

In this episode of From Startup to Exit, we sit down with Ezhilarasan (Ez) Natarajan, CEO and co-founder of CoreStack, to explore the company's remarkable journey from startup to global leader in multi-cloud and AI governance. Ez shares how his years building Microsoft Azure inspired him to tackle one of the biggest challenges facing enterprises: simplifying cloud adoption. He discusses how the team validated an idea that investors initially dismissed, built an enterprise platform from the ground up, secured early design partners, and scaled CoreStack into a company serving thousands of customers worldwide. The conversation then turns to the AI revolution, where Ez offers a compelling vision for the future of AI governance, enterprise adoption, and the transformative impact AI will have across every industry over the coming decade.

In this episode, you'll hear about:
- How Ez Natarajan founded CoreStack after recognizing the growing complexity of multi-cloud adoption and why the market initially failed to see the opportunity.
- The startup journey—from self-funding and building the first product with design partners to securing product-market fit and raising venture capital.
- The lessons learned in building an enterprise software company by listening to customers instead of following conventional investor wisdom.
- How CoreStack evolved into a leading platform for cloud governance, helping enterprises reduce costs, improve operations, and strengthen security.
- Why today's AI revolution is fundamentally different from previous technology waves and what it means for businesses over the next decade.
- The emerging importance of AI governance—including managing AI costs, securing enterprise data, governing AI models, and controlling AI agents.
- Leadership insights on building culture, attracting exceptional talent, integrating acquisitions, and scaling a mission-driven technology company.

Brought to you by TiE Seattle
Hosts: Shirish Nadkarni and Gowri Shankar
Producers: Minee Verma and Eesha Jain
YouTube Channel: https://www.youtube.com/@fromstartuptoexitpodcast

Transcript

SPEAKER_00 0:01

Welcome to the show. Today we've just found this video and co-founder, of course, uh support companies from Mark Walter and startup, global leader and multi-cloud. And I thought we'd see Microsoft and Marco one of the biggest clouds, uh simple flying uh, when you look at hard findings, podcasts like Thai Seattle and Thai global non-cost and that's focuses on fostering entrepreneurs. Thai Seattle offers a range of programs, including technology panels, the Thai Entrepreneur Institute, and the Thai Seattle Angel Program. We encourage you to become a TIE member so you can gain access to these programs. To become a member, visit www.seattle.ti.org.

SPEAKER_04 1:39

Hello everybody. Welcome to another great episode of uh From Startup to Exit. Shirich Carney and I are very excited to be uh recording this one. This is very close to us, both in location and our interest. We are happy to have Earl Arasan, as he is commonly known amongst us as Easy. Earl Arrasan Natrajan, CEO of Core Stack. To introduce him and uh kick off the podcast, I'll hand it over to Shirish. Sharish?

SPEAKER_05 2:10

Yeah, thank you, Gavri. So uh again, we are very pleased to uh welcome uh EZ Natrajan, who's the CEO of Core Stack. Core Stack is a leader in multi-cloud and AI governance. And uh Easy is also a uh Thai charter member. So thank you, uh Easy, for you know, uh, your time and mentorship for other folks in the Seattle community. But they have raised over $85 million. Uh so uh they're a very mature company, uh startup company. And today we'd like to explore kind of their journey with uh Core Stack. So welcome, Easy.

SPEAKER_01 2:45

Thank you, Gauri. Thank you, Sharish. Uh happy to be here in your uh Startup Exit Pack podcast. Uh, thanks for having me.

SPEAKER_02 2:52

Uh looking forward to sharing our journey and uh what we have uh ahead of us.

SPEAKER_05 2:57

Great. So before we get started, that would be great if you could uh describe a little bit of your background uh before CoreStack, what were you doing? And and then we'd like to explore kind of your journey with the company.

SPEAKER_01 3:08

You are kind of revealing my age, uh Sharish. So by saying that I'm in the industry for like 28 to 30 years, uh, you know, I I don't feel young in the AI world, but it's been a 30 years journey in the technology world.

SPEAKER_02 3:21

Uh graduated uh in the late 90s uh from the University of Madras and Chennai as a computer science major, uh university and campus school modelist, worked for captive uh uh RD centers in India from Alcatel to uh I2 technologies to Texas Instruments for the first decade. Use that uh you know uh experiences of putting technology for larger captives RD transformation in the first decade to uh building larger impact uh through services uh arms. Um, work for a company called MineTree. Eventually that helped me land in Microsoft, helping uh build Azure through a partner ecosystem for the uh late 2000s to early 2010s. I've been part of uh building Microsoft Azure in China, Australia, building the Southern Cloud uh foundations uh for Microsoft, which led me to come and land at CoreStack, building a platform, an intelligent platform that would help customers in the digital journey. So the first two decades, technically in the corporate world, uh building large-scale technologies and platforms for the world. And the last decade is uh Core Stack journey from inception to where we are today.

SPEAKER_05 4:33

Great, great. So let's talk about CoreStack now. Uh how did the idea of Core Stack come about? And uh how did you vet the idea?

SPEAKER_02 4:45

Interesting, uh interesting question, right? Uh you know, it takes me back to almost a decade, uh, 10, 10, 11 years ago, when we are really looking at what is the problem we want to solve. And these are the days of uh, you know, for Microsoft, we are building China, the Blue Cloud Services, for Australia we are building OpTSES Cloud. That's when the idea of Costa came in. My co-founders were doing private cloud capability creation for large companies, you know, as a separate individual consulting small business, sitting out of India and sitting in the US, I was helping uh you know, Cisco and Microsoft and Optus to come together on one side, uh, take Microsoft to China on the other side, uh, and then there was Dell private clouds, HP private cloud, fidget source cloud, all this were machuming at that point of time. Uh, while the cloud players back in 2015, if you really look at it as AWS, Azure, IBM, and Google was probably not in the foray, then there were a lot of uh you know, VMware and hypervisor-based cloud providers. All the botch products were getting converted into as a service consumable technology, and we saw just two different worlds of living. Today, the words of data center is becoming more prevalent. Ten years ago, people were moving from data center to cloud, and the technology consumption was uh complex. The CIOs, CFOs, and CTOs were having tough times in putting this new piece of technology meaningful for companies to create the outcomes. And that's when the idea of CoreStack was born. All the cloud providers were cash negative, customers want to adopt technology, but there is a complexity between them. Between the providers and the consumers, there is a complexity. Can we simplify the complexity by creating an intelligent platform? Way back to while today is all about AI, we called the platform as an intelligent platform that helps customers in the technology adoption. That's how PostAt was born. So happy to answer more questions when you ask me the dimensions of the nuances that are coming in. But it's it's a well-thought uh you know entry into the cloud world of making technology easier for customers.

SPEAKER_05 7:02

Got it. Okay. So how did you uh vet the idea? Did you uh was it based on your experience, you know, working with different cloud providers, or did you talk to CIOs, or how did you know that you know what you're trying to do made sense for the CIOs?

SPEAKER_02 7:17

When we had the idea, the idea came from observing the customers, not talking to them. We were just observing the customers having challenges. I see. They were talking to their technology partners, this could be the you know, SI partners or their technology consulting partners. It could be the Big Four or the Big 20 we've been working with. Uh, they had an in-house team that that was building methods of living in a traditional world, building the new age, coexisting between the two. So we saw uh the challenges uh when they adopted, you know, some CIOs, you know, between 2015 and 16, 17, if you really look at it, Gartner was publishing some data which was very uh uh you know very interesting to look at. The C-suite change in Fortune thousand companies between this 2015-2017 period was almost 70 percentage. Meaning if a company does something on cloud and they were not able to do it well enough, then it backfires either on the cost front or the security data compromise front. You know, we have seen companies from T-Mobile to Primera to many companies that that were getting compromised on the data side because they moved to cloud, they couldn't control, they couldn't manage. And the companies like Alaska Airlines at that point of time, they moved significantly to cloud. Their monthly bills have started shooting up in tens of millions of dollars. That was not their technical IT spend at the point of time. So we knew the problem existed. We created this idea of solving this, simplifying it, providing a method so that the technology can be meaningful in terms of cost, in terms of security compliance, and in terms of operations, so that total cost of ownership is reduced. But when we went with an idea to bounce it off with some of the you know uh market leaders from analyst standpoint, uh advisors or even investors back in 2015-2016, not many people believed in our idea. Because we said we want to solve digital adoption problem across multiple clouds, and we want to solve it for the CIO, CFO, and CTO. When we said this statement, the acceptance level, because they thought we are boiling the ocean. First of all, you are doing multiple personas, you are trying to solve the problem, and you are trying to solve this multiple persona problem for multiple clouds, which was not even accepted as a problem. Multi-cloud was not even accepted as a norm back then. So when when there was lack of acceptance from investor segment, from advisors, from the analysts, then we had to really look back our customers. My philosophy that we used at that point of time was for people who know uh Mahabharata, right? Uh, they know this uh story where uh Dronacharya takes all the 105 kids when the Gauravas and Pandavas all together uh with a bow and arrow and asks from Duryodhana, the oldest one, to everybody, right? Uh 104 kids, he says, Hey, look at the tree branch and see what you see. All the kids see the branches, the leaves, the stems, the you know, all of that stuff. No kid sees there's a bird with a bird, and there is an eye of the bird that is visible. It's Arjuna who comes and says, Oh, I can see the bird. Then Dronacharya says, Sure. Then all the 104 kids come and see again, hey, what do you see? They all see, yeah, there was a bird, yes, we could see that. Now we felt that we were in that kind of a moment. Not that we were prizing ourselves that we were Arjunas. We were, you know, the two people at that point of time knew there was a bird. One, there was Dronacharya, who knew there was a bird because he was asking the question, and second, there was a kid who could see that. There was all other people who could not see that. And uh we started assuming that if we were able to see the bird, who's our who's the dronacharya here? It was not the investors, it was not the VZs, it was not the advisors, it was not the market analysts, the dronacharya was a customer. So instead of observing the customer, we went and started talking to the customers, and they told us, yes, if you have something of that kind, we will immediately use it.

SPEAKER_06 11:29

Right.

SPEAKER_02 11:29

And we said, okay, so there is a bird, and customers are at least confirming it. And we started working on building the first version of uh CoreStack. Between 2015 and 2016, we were in stealth mode. We didn't raise any money. We said, let's build the platform and go give it to a customer and let the customer tell us is it worth building to the next phase or not? So it took about six, seven quarters for us to build the uh original platform from scratch. We were conscious of doing it without any open source, without any you know, uh third-party IP. We wanted to build the IP uh which will differentiate us for the next decade. I'll talk about some of the uh you know uh fine fingerprints that we have created which made us today meaningful in the AA world. Uh but that's how our journey started.

SPEAKER_05 12:20

Right, right, right. So uh we'll come back to the kind of uh what the solution looks like initially. But curious, uh, you know, how did the uh team come together? Who are your co-founders and how do you decide that you wanted to work with each other?

SPEAKER_02 12:34

So we had four of us as a co-founding team coming together. In fact, uh probably uh a couple of more who joined us from day one, but four of us had to because nobody was willing to put any money and we had to create a product from scratch. So there has to be funding that is required. So four of us uh happened to contribute at the initial seeding as as a you know founder-funded uh company. Four of us came together. We were all aligned just on the vision. The technology is fast evolving. Customers want to use the technology, but there are challenges, there are complexities. Easing that complexity will create an opportunity for us was the vision. And all four of us were well aligned. As I said, three of my other co-founders, uh Sabah, who's my CTO today, Krishna Kumar, we call him KK, who's our COO, and and the other co-founder, Thiru, who's no longer working for us uh due to a family situation during COVID, he had to step out. He lost some of his family members.

SPEAKER_06 13:33

My God.

SPEAKER_02 13:34

So, I mean, just all families went through in COVID, so it is an unfortunate event. He is invested, he continues to support us, but not actively working for us. But between Sabah K and Thru, they were running private cloud creation consulting company between 2008 to 2014. It's a boutique, uh private data center, private cloud creating company on OpenStack and open source standards. Uh, they were running a few million dollar revenue generating business, uh, sitting in India, earning a few hundred thousand to half a million dollars per year was a great uh money. Uh, but when we came together with this vision of creating an intelligent platform, we decided as a collective team to shut down that business what they were running in, and the profits that came in became their investment. I have gone on a stealth mode of bringing additional money required to build the foundational platform. That's how we four came together. And with the vision in the 10 years, my entire leadership team, even today, we have people from Microsoft to Amazon to Google to you name a large player, my leadership team is comprised of that now. People are here just for the vision and the impact we can create. That is the one the strongest element that is keeping us together uh for the 10 years from where we were to where we are today.

SPEAKER_05 14:53

Right. And did you know these folks? Uh had you worked with them in the past?

SPEAKER_02 14:58

Uh yes, uh, we have worked uh you know, prior to that few years, uh, you know, when we were going to the Opta C S mode and then Opta C S Cloud. There were working uh relationships, uh more as a you know, first party, third party kind of relationship. But then you start uh discussing what are you doing, what do you see the future of uh you know business, technology, what do you want to do, and what I was doing versus what I'm anticipating in doing. It's it's uh think of it like dating for at least a couple of years.

SPEAKER_06 15:32

Okay.

SPEAKER_02 15:32

We've been dating for more than a couple of years on this idea, brainstorming. And 2015 is when uh we decided, okay, let's go start it. And 2016 we could give a uh structure to it, and that structure became Core Stack.

SPEAKER_05 15:49

Right. Got it. So what what what did the first version of the product look like? What are some of the uh obviously, you know, uh you could boil the ocean, as you said, but you have to focus on a few things. Uh what did you decide to focus on?

SPEAKER_02 16:01

Well, the name Core Stack itself, you know, Core Stand stands for, you know, uh, you know, you can call it as a uh combined orchestration engine. So we started off with multi-cloud orchestration, maybe a complex word for non-technology people. Basically, you know, how do you provision the cloud capability services, whether it is a server, whether it is a VM, whether it is a storage, whether it is a database, whether it is a network service, how do you provision it and make it operating as if you were in the traditional world? So your standard operating procedures, your workflows, your methods, you know, your structures, whatever you did, it can be easily applied in the cloud world. Right? So simplifying orchestration for that purpose. And we started off with uh you know cloud ops, what today we call it as, right? Cloud operations, provisioning, orchestrating, managing, making it simplified, as if the traditional world and the new world is Apple to Apple. They're on Orange and Apple. That's where we started. And the first version of it we offered to uh some of our earliest customers, the biggest telecom brands in the world, biggest network providers in the world. Uh, when we gave it to them, within 30-60 days, I mean they were like design partners for us. We didn't sell the product to them. Through the network of references we had, I know, hey, can you try this and tell us what is the value that you see? Within 30-60 days, this some of these big companies have come back and told us we see efficiency increase going up more than 60 to 80 percentage, or the productivity increase. At that point of time, cloud was used as a spillover capacity because everybody had a data center, and they were using cloud as a spillover capacity. One of our earliest customers said their spillover capacity was more than quarter million dollars per month, and after using our product, suddenly they could save more than $100,000 per month. This was like significant. This was not and not even anticipated. So their feedback to us is can you make this enterprise great? Can you make it a large scalable product rather than a point solution? Start solving the other pieces that we shared in the vision. Hey, we'll solve your financials, we'll solve your operations, we'll solve your data and security side of it. But we gave only the operations side of it. But with operations, without even we creating the financial savings, they started seeing the benefits because operations helped in indirect savings. And they measured it, we did not measure it. So that's when the FinOps side of it, the SecOps side of it, the App SecOps, it's all a natural extension that our customers got drove to the next level. But yeah, the first version of it, we started off with CloudOps, eventually with FinOps.

SPEAKER_05 18:44

Right. So uh how many design partners did you have initially?

SPEAKER_02 18:48

Uh we should have had probably about eight to ten. Uh okay, that's a lot. That's a lot. Oh, eight to ten in telecom domain, RD, healthcare. Education was probably three or four. Uh in fact, at some point of time, the first, second, third year, uh before our funding happened, the domain of education even became uh uh a fast mover uh because we were able to provide next generation learning, uh you know, stop all the rogue usage that were happening in the education domain. Because if you look back 2016-17 is when today's AI is precursor of blockchain and Bitcoin mining was big time. Uh today Nvidia is uh you know a five trillion dollar company because of AI. Back then, Nvidia was seeing a significant uptick, not because of AI, but because of uh blockchain mining. So all the training institutes, education institutes, whoever has this GPU, or if they have their traditional compute, the students were going and creating mining algorithms, running it on a free compute, and the companies were really coming and asking somebody has to create a god rail and stop this mining thing for which we are ending up paying. So we saw an uptick on education, but primarily the largest use case was avoid or stop the rogue rogue use cases uh as a cloud operations god rail, and and eventually we moved into more fortune, you know, thousand, fortune, two thousand customers. But for a short period of time during the mining days, we became the godrail of many companies, including Great Learning, Simply Learn, you know, uh many of the large education companies and some of the large universities uh in India. But you also had corporate uh clients, you said telecom and others who were so we we had Fortune 500 telecom customers, networking device customers, healthcare companies in the US. So they were our earliest uh design partner.

SPEAKER_05 20:46

Right, right, right. And so you felt that based on the feedback, they were saying they were saving several hundred thousand dollars a month that you had product market fit with your solutions.

SPEAKER_02 20:56

Right, yes. When we took the customer feedback and started talking to the uh analysts and advisors and VCs, then the reception reception started improving or increasing. Okay, that there seems to be a problem. You're solving it.

SPEAKER_06 21:15

Yeah, yeah. Yeah.

SPEAKER_05 21:16

Yeah. So uh at what point did you go and uh then uh you said initially the VCs didn't want to give you any money because they didn't believe in your the problem you're trying to solve. Now that you had you know 10 design partners who are all saying, you know, we're saving money, large amount of money, and we want you to do some more stuff. Did you have a good story to go talk to the uh investors?

SPEAKER_02 21:36

Oh, uh it's it's more than an interesting story now if you have to recall uh Sharish. One of our uh customers, the design partner customers, uh, when we did not end up selling for them, the first largest telecom company that we're working with, they came and told us we are willing to pay for you to add this capabilities. So they gave the first revenue check for us uh in 20. Uh 2016 itself, end of 2016. Uh, they gave the uh check. Probably a smaller check, but then it was not a sales check. You have a design partner coming and telling you, here is the money I'm willing to pay you. That that occurrence continued out of our eight, ten design partners. Three of them have been willing to pay. And in fact, one of them in the healthcare space, a service provider who's helped serving a healthcare customer, came and gave a quarter million dollar check, not as revenue, but as an investment. No questions asked. He said, I know you guys are going to take money at some point of time. You are going to be meaningful in the future because they implemented then exited version of CoreStack for one of their healthcare customers and they saw the benefits. And a customer was pricing and their business from the customer was increasing because of the way they could do it. It's a body healthcare provider, you know, we can call them as a managed service provider back then. They gave us a quarter million dollar check even without we asking for money. And we ended up using the check probably after six months. Uh, and that was a trigger for us to say that okay, looks like we can go and tell the story to the investors now, and I have the first quarter million dollar check uh that is available from my customer, right, who's using the platform. Right. And uh 2017 is when we uh when I started again approaching the investors after 20 uh 15 and 2016, early when we tried almost next 18 months, we didn't go for fundraise. Again, mid-2017 to late 2017, and they started approaching the investor investors for investments. I had a check already.

SPEAKER_05 23:37

Right, right, that's great. So so was the reception the second time, was the reception a lot easier, a lot better?

SPEAKER_02 23:44

Oh, lot a lot easier. By the time, you know, by 2017, mid-2017, AWS became cash positive for the first time at 4 billion revenue or so. Every other cloud provider was still cash negative, including Azure, including IBM. Google was just starting to make inroads. Oracle was talking about the OCI first version of it. They're all cash negative, but AWS became cash positive. Uh and multi-cloud becoming a norm. Customers were talking about you know, we are not going to just put all our workloads in one cloud technology itself. We are going to get the best of multiple providers. Yeah. You know, 18 months fast forward or 24 months fast forward between my first iteration of fundraise and the second iteration of fundraise, the market changed. The customers spoke. There has been an increased acceptance of multi-cloud as the terminology. And we were the first one to say here is a platform that simplifies the technology adoption for multiple clouds at once. And we also created certain differentiation. Uh, for example, we don't have an abstraction layer. CoreStack uh is unique in extract in exploiting the power of each technology by natively working with the cloud provider's standards and not creating abstraction layer. Tools, even a point solution, a migration tool, a FinOps tool, a security tool today. Even this, if you had to consider it as a security point solution tool, they all have an abstraction layer. ColdStack was and probably continuing to be the only one that works with every cloud, every AI in the native way without having an abstraction layer. That also differentiated us to be more accepted by investors.

SPEAKER_05 25:35

So in your uh first external round, how much did you raise and how who are your investors?

SPEAKER_02 25:41

Uh the first round of investment uh, you know, uh we had probably five or six term sheets uh in 2017 end, 2018 beginning is when we have taken the money. I think we raised about uh three and a half million dollars, uh and one and a half million dollars came from uh Z Fi Capital.

SPEAKER_05 26:01

Yeah, just yeah, Z Fi Capital is the in a fund that a number of Thai charter members.

SPEAKER_02 26:08

ZFI was you know thanks to Thai Seattle for creating the first of its ecosystem of creating an investment opportunity through a you know layered structure that the Thai Seattle members have believed in in the technology we are creating, thanks to all of them, and Z5, which did a wonderful diligence on us from uh Bay Area. So it was a three and a half billion dollar round, I believe a portion of it was uh by Z5, led by Z5 from valuation standpoint. But there are multiple uh other HMIs and angels and smaller investors who participated in the round.

SPEAKER_05 26:41

All right. Uh uh over to you, Gabri. Yeah.

SPEAKER_04 26:44

So easy. This is a very fascinating conversation. That's kind of shifted today, right? And so everything is AI today, right? There's both uh fear and anxiety, uh, excitement and opportunity. Both are you know colliding every morning, whether it's be enterprise or individuals, right? You have a unique story to tell, both as an enterprise, uh, both as a company serving the enterprises, but more importantly, as an entrepreneur, how did you how are you looking at it and how are you going to address it? Because this is not a simple thing. It's not like I can flip a feature switch and you have to now rethink this thing. Looks like you had phenomenal training, people not believing in you. So you gotta do this all over again. So how how are you thinking about it? What your your thoughts and company's approach? Can you kind of elaborate on both of those?

SPEAKER_02 27:42

Uh definitely, Gaudi. Uh very involved question. I may have to simplify it as much as possible and tell this in a way of what is driving us to be meaningful in this AI world, right? So, first of all, uh in my opinion, 90% of the people in the technology world does not understand today's AI. Uh, they go back to you know the data science version of the last 10 decades of all the methods in which the technology industry was trying to create AI and the neural network languages that we have been talking about since the 80s, ever since the neural capabilities and derived AI was getting created, versus what we are seeing as the real AI today. While the output created is the AI output, the technology beneath is not the same. Today's AI is possible because of the massively parallel computing capabilities that are available through the NVIDIA and the hardware uh capabilities that are available. Uh today an NVIDIA processor can process about uh 13 terabytes of data per second. Latest Nvidia processor can process about 50 trillion transactions per second. And the next year's one, we are anticipating it that it will create 200 trillion transactions per second. This is what is powering the AI. Uh, you know, the models are the newer version of the neural uh you know logics that are being built in the in the way it is able to process tons of data in a fraction of a second and find a needle in the haystack. Right? So I want to lay down this foundation of AI. And what it can do for a larger civilization is what we are thinking through. Because today we have 2,000 plus customers. Every third brand in America probably is our customer already. Whether it's your retail brand, whether you're flying on the flight, whether it is American defense, they're all customers of Costa. So we have gotten the right to support them, to right to enable them using the technology from Cloud World. So when cloud is transforming to AI, it's very important for us to understand what is the driver, what is the foundation, and what are the challenges that are going to come in and how to solve. So, as an entrepreneur, what we are seeing is how these technologies are going to impact uh the larger civilization itself. Each industry is going to explode uh more than 10 times in terms of innovation, in terms of how the industry has traveled. Uh, you know, in the last 50 hundred years, I see uh a three-dimensional shift up to 1900. Uh, you know, how uh everything functioned in the world prior to the invention of electricity, prior to the invention of uh you know uh aeronautics or aeroplanes, uh capability for humans, prior to the invention of uh you know computers as such. Then you come to 1900 to 2025 with internet, with computing, you know, mobile, available to everybody, you know, whether it is travel, healthcare, uh, you know, manufacturing, whether it is food industry, how they all changed significantly from 1900 to 2025. We are sitting at a cusp of AI which is going to give 10 to 100x boost for every single thing humans are interacting. And we are touching all these customers, all the companies that are enabling in this. So we need to empower them in using the today's AI capability, which is massively parallel capability of processing unimaginable amounts of data and giving velocity to the business to do things better, efficient, faster at a reduced cost. When all of this happens, there are you know drivers. Today's AI, while it is super fast, it's expensive. Expensive because of energy, because of the rare materials that are not available to create enough hardware capability. So there needs to be an orchestration or usage of the capability in a meaningful way. Next, when you start putting in unimaginable amount of data, the privacy, the compliance, the security, can it what can be done to this data is everybody's fear. Today, when Anthropic says that I will not open up Anthropic uncontrolled to even American defense, you know, they're having a larger argument with American defense because the power of AI today, nobody knows, including the creator. It's not the Amadays of the world, it's not the Sam Albums of the world. They are cautioning, they are warning that hey, we have created uh a technology without understanding the implications of it, because it's a model, it's a data. You can, you know, it can be unleashing unimaginable things if it is not controlled. So the customers who have been using CoreStack for their cloud governance, hey, create the guard rays, the use cases that I talked about 10 years ago, the education institutions coming and say, hey, I'm I'm giving a learning opportunity for a student, but that student is going and doing a Bitcoin mining, you know, help me put a card ray. It's the simplest of the use case. I think 10 years later, now customers are coming back and saying the same thing. I've created a model, I have AI capability offered by the cloud providers, or I can even buy it from Nvidia today. I don't know what the model is doing. I don't know how to control it. I don't know what I am even compromised. When will I even know it? Okay, you know, that's one side of the problem. But the other side of the problem is I'm starting to spend, you know, hundreds of thousands of dollars quickly moving to millions of dollars. Some of our large customers are saying their AI spend uh six months ago was a couple of million dollars. Just in six months, it has gone to twenty to thirty million dollars, right? On on an annual measurement basis. It's a 10x growth in the cloud spend AI spend by customers, and okay, something you must be creating meaningfully, great. They don't even have a clue of what's going on in AI. Who's using it? Why are they using it for? What are they using it for? Can it be done by one AI versus other AI in a better efficient way? And during this process, what is being compromised? And what do I need to spend tomorrow to give a meaningful outcome? There is no RYA measurement, there is no TCO measurement, there is no control godrail. This is what customers are pushing and asking us for. And we are taking a meaningful approach of how the civilization is going to move forward and what is that we need to build as a layer for the technology, for each foundation to be meaningful. So we are thinking through three layers of uh you know creating uh a governance layer for AI. One at the token cost level, bring visibility, control, management ability for companies or even individuals when they use AI. What is it, what is that they're using for, why they need to use, what is that they can use alternatively to get the same outcome delivered with a lower unit of cost? Then get to the model. Are the models doing the right thing? Are they trained in the right way? Are they governed from not doing any unethical stuff, anything that accidentally be allowed by a persona to create an outcome that is not acceptable? So those are the governance structures that need to be put on the model level. Then comes the agents. Today, every software company in the world are starting to offer agents, whether it is cloud providers or whether it is the application providers like the sales forces of the world, SAPs, Oracles, everybody is providing agents. Customers with cloud today, they're building their own custom agents. So there's expected to be a lot of ungoverned, bought agents that are working on model that nobody knows what the model will do if it is not controlled, using tokens for which you don't have a control. And this is the new generation problem. And the advantage of it, if it is used perfectly, what I foresee as an outcome as a sample, right? Today, point A to point B, anywhere on earth, humans need probably about 24 to 36 hours. I foresee in the next 10 years AI will democratize this to bringing it less than four hours. Point A to point B, anywhere on earth, humans will be able to travel for less than four hours. Everybody on Earth will have food and shelter. That's one of the predictions that we anticipate because we have customers like teleforms who's producing the, you know, it's an $8 billion fresh uh fresh-produced company uh that's working across the globe, having fresh-produced forms across the globe. When we enable them for cloud and AI, the transformation that they are anticipating is put food on the table for every single human being on Earth. So if this kind of transformation has to happen, the technology must be allowed to do ultra fast things, but with a great God rail where we can believe that only good things will happen. It will not lead to bad things. So we are at a juncture of assuming the governance control plane for AI, being inside of the technology and being outside of the technology. We need to play the intrinsic and extrinsic role. So it's it's a larger impact and responsibility given to us, and we have earned the right managing $40 billion plus per minute. As we are talking today, CoreStrack platform manages about $40 billion of cloud and AI being used across our customers. And taking it to a few hundreds of billions of dollars, we have the right to win in the space to create the God rails, governance structures for AI.

SPEAKER_04 37:55

So this leads to you talked about from inside and you having the you know, you earn the right, so you're also inside. Right. Now, the important question as a leader, you have to, as you take your employees along this journey that you're articulating. There's a culture shift. There is uh there is a shift in thinking, how they should react, etc. And you specifically, CoreStack has also grown through acquisition, which means they have you have others coming in with your own culture. So for our audience, especially entrepreneurs out there, the question of culture, because you have those who joined you from the day one and those who joined you recently through an acquisition, you've got to bring everybody on the same table so that your position of being inside is well respected as you go forward. Because at the end of the day, your people have to implement what exactly you're articulating. So how are you uh bringing that together for uh Core Stack, the larger Core Stack, all your brands underneath uh that you have?

SPEAKER_02 39:07

Thank you for bringing that dimension. Yes, we did have uh meaningful acquisitions in the last few years. Uh as you were growing and getting access to certain segments of the market, including the uh US federal government access to structure where we have made an acquisition. Most recently we have acquired BetterCloud. This company was doing what we were doing on the infrastructure layer for the infrastructure cloud. Better cloud was doing it at the software layer. All the applications typically the customers are using provide the spend visibility, provide the automation, provide the data security governance on the SaaS layer. Today in the AI world, from infrastructure to data to the models to the SaaS, you need the combined visibility. So this acquisition was well thought out in terms of bringing meaningful understanding, a single source of truth of technology platforms that are going to contribute to AI to create the velocity. So today, it's a velocity game. Everybody wants to move fast. Today, one of the internal metrics of every company is how much of code is written by humans versus how much of code is being written by machines. And I foresee that in the next couple of years, the entire world is going to run probably with 80% of machine-written codes. Uh, there's going to be only 20% of human-written codes ever. In that mode, what is critically important is underlying data of how the machines can build the experiences, the workflows, the integrations, you know, whether it is mobile arts, whether it is SaaS experience, whether it is uh you know IoT experiences that we are going to have. So the intricate piece that was missing for us was uh the SaaS experience layer and the data that are being used. How the last decade of companies that have used the technology, right? So we had the opportunity to go create it, or we had the opportunity to go acquire something and fast integrate. We thought velocity is the name of the game, so we don't have the time to go recreate it. And the people who created it, they were also finding it difficult because in the new world, platform is the new critically needed element rather than a point solution. You know, every platform company that exists today in the AI world, you know, this is not my belief, this is this is the larger belief in the technology world. Every platform is going to be thriving because they are needed for AI to function credibly well. And we created a platform, but we were missing certain components. We brought them in. And now the the next thing is bringing the people together, creating the culture that we can continue to innovate faster, better, and creating the outcome, laid out processes from a people standpoint, um, you know, organizational structure standpoint. Things that are connecting us together is the level of transparency we built to the company, the level of vision, and the accountability we empower with our people. Here is the vision, this is what we want to do. Here are the challenges. No one has solved it. We have the right to solve, and some of our customers are already solving, and allow that to be imbibed by every single individual in the company. Today we are a company of collectively 250 plus people. We are looking at, hey, we don't want to expand this headcount in an in an unpredictable way, just from a growth standpoint. We know that bots are going to be counted as employees eventually, but we have empowered the existing employees to use the technology to multiply their capability into 2x or into 3x. Thereby, we are not one of the companies going and doing a significant layoffs as part of the AI innovation curve driving them. So it's a responsible growth, responsibility of value creation given to our people, and the power of using the AI to create a meaningful AI is given to our people. In all of this, we are being very transparent. And we let them hear from the customers. When the customers are coming in and telling us that the 2000-plus customers that we have, they are crying loud in terms of what is needed for them, what they needed yesterday, what they need today, what they need tomorrow. It's loud and clear on the walk. And the success is attached to them to create that, you know, what we have done over the last 10 years, we have given 20% of our company to our employees. So I'm I'm proud to say that. I don't know how many companies out there would have done it. Uh, this is another cultural aspect of not just giving the responsibility, giving the roles and titles, giving them the accountability, go do it. Rather, when we create the great outcome, when we create wealth. Everybody shares it in a meaningful way, not in a way that the investors or the original founders take disproportionate outcome, and the people who really contributed for the outcome do not make really much. So we have imbibed certain principles and we stand by it. This sends a message: I'm not here as an employee. I'm really creating a dent in the world, and the company res. And the company is not meaningfully coming and laying off us. Before hiring, they are very thoughtful, uh very mindful, and uh you know they are not unmindful of laying off. So this message has repeatedly gone in to create. You know, when we have new employees come in, all we do is open up the floodgate and ask all the employees to collaborate without control. Uh nobody can go convince the incoming employee with certain level of mindsets through a management or through an organized structure, right? While the organized structures are there, people, when they talk to existing people, hey, why are you here? You've been here for a long time, you've been creating. Why am I here? The purpose, the value, the culture of transparency, accountability, and stakeholdership makes them believe instantly, okay, we are at the right place to solve the right problem. So I know it's a little longer answer, Gauri, but this is the philosophy that we have followed.

SPEAKER_04 45:26

No, I think that's that's a very insightful one, uh, especially for audience, because you know, you they you get lulled into the belief that in the new world order of AI, we can just do it all with bots with less number of people, which is a very important message. Because the anxiety that is being created among the among the employees is because they don't understand the path. These are all very smart, highly skilled, knowledge workers. So they're able to handle that. Right? So that's kind of where your philosophy fits in with the with all the entrepreneurs that are starting. More importantly, because you've been on this 10-year journey. You start as a cloud native company, now you're becoming an AI native company, and there's lots of lots of companies like you there who are in the cusp of that change. I think if the clarity you provided was extremely good. Let me circle back to the one thing that you said in the beginning, right? It was clear to you there was a bird in the tree, right? You weren't sure if the others were seeing it. Others were telling you we don't see it, right? Now your customers validated that bird and you said, okay, let's go, let's go uh, you know, go after that one. So that was that was that was excellent. As you grow, right, your customers are both pushing and pulling you all the time, pushing you to give them more, pulling you because they want something different than what you're offering. So the bird is still there. How does as a mature company now, as in mature, as in the journey you've been, not not in time? As a mature company now, how are you uh viewing the bird? Because people already believe you. You have investors, you have employees, you have customers, you have partners. The circle is complete, but you still got to look at the bird because it's you still have to catch it. So how are you approaching it now versus when you started the company Corsak?

SPEAKER_02 47:29

That's a very interesting and wonderful question, uh, Glaulia. As you're asking, I was trying to uh create a metaphor that I can convey. Uh it's like the flock of birds that we suddenly see in some parts of the world, right? Where when you go to the uh snowy areas, you know, you you see not just one, hundreds or thousands of birds flock together and suddenly they create a nice uh you know experience on the sky. Maybe when we are on vacation, you would have seen it, or even in uh some of the movies like uh superstars, you know, 2.0, you would have seen uh how they were going behind the telecom world affecting and all of that, right? So it's it's like today it's not one bird. With AI, the opportunity is a flock of birds, hundreds and thousands of birds are sitting in. There's nobody questioning whether there's a bird or not. It's a question of how do we manage this? Because there can be a menace that can be created. Everybody is trying hard, starting from making these thousands of birds go to millions by solving the energy problem. Today, Microsoft says that we have hundreds of thousands of uh Nvidia computes sitting in our storage yards because there's no power to power them. The world does not have enough energy to power the capacity. The more energy problems get resolved and the unit energy cost goes down, the abundance of AI is going to suddenly increase because the unit cost of consuming AI will reduce and everybody will it's starting to get democratized, everybody will start using more AI. When we start using more AI, once the cost problems reduce, which I foresee in the next few years it will naturally happen as more innovations on the energy and the radar mining and the manufacturing sides get solved, then it's going to be more and more responsible AI, ethical AI. You know, models need to mature significantly better. But during this time, we're also going to experience artificial superintelligence is going to be born. Meaning today, AI is a bit lower in capacity for a best human thinking. Right? The Gen AI part of it. When AGI comes in, the artificial general intelligence comes in, AI is on par with the best IQ human being on earth, and it is available to everybody. When ASI comes in, which I foresee in the next five to ten years, when it happens, it's going to be better than humans. And that is the biggest of the problems. During this period of next five years, we should also anticipate quantum is going to hit reality. Quantum research is a big time by majority of players, the big time players. And when quantum plus AI combination is going to be detrimental, it's going to accelerate the pace of innovation, it's going to disrupt cybersecurity as a whole domain, right? Every single password that has been created in the world can be deciphered, can be cracked by one quantum computer in less than two hours. So we are anticipating some of the innovation, the velocity increases and the change in dimensions and the catastrophic outcomes that are going to come in. And because we think as a God rail provider, we think of it as a governance provider to make sure that this AI is meaningful for every company. So they are able to create the outcome faster for their customers. So we are at a juncture where we need to deliver what is needed today and deliver for tomorrow. This is why I'm saying there are thousands of birds, and some of these birds may be maniac in trying to fly in the river direction or hurt each other or create outcomes that are less desirable. And one such major event can catastrophically create a trillion dollar impact on the stock market today. While we are all excited about markets moving up by trillion dollars, we think that catastrophe can be created humanly today by powerful people. The same catastrophe can be created by one misuse of technology or wrong use of technology, which we will all see in our own experiences in the next couple of years. But then the world's prerogative will be to not have that repeating. It is a problem of anticipating the bad element, the bad character, the bad behavior, and preventing it rather than reactively solving it. Whoever is going to move the needle in a preventable mode is going to be the winner, while the the ecosystem itself is going to be democratized cheaper for everybody to adopt it. And today's the bird is a preventable, ethical AI capabilities that can empower everybody, and we are putting our foundations for that.

SPEAKER_04 52:37

That's pretty phenomenal. We could go go on. We should have a two-part episode, Sherish, with easy. We have to go on for a long time on many more topics. I've not even gone around to talking with you. But easy, we Sherish and I share this very proud moment that Costac, born through the Thai Seattle ecosystem, has uh flourished into a major force not only in Seattle, but in the world order of enterprises. And uh you also have government customers, and uh we're very, very happy for to see one of our own, you know, uh lead the charge on entrepreneurship. So we wish you the very best and uh we hope to uh continue to support your endeavors as uh as TICL. So thank you very much, uh Sharish. Back to you.

SPEAKER_05 53:32

All right, thank you, Easy. Uh great conversation. We'd love to have you back at some point, especially to explore the AI governance piece in more details. But thanks for the conversation today.

SPEAKER_02 53:41

Well, thank you, Sharish and Gauri. Really appreciate it. And thanks to you know, TICRL team and uh the ecosystem as TICRL created for remaining Thai groups to follow as an investment arm. Super proud, super thankful to all of them who believed in us in our earliest days when the world did not believe us, and uh glad to be here creating a bigger dent, uh meaningful dent for the world. You know, looking forward to uh days of uh collaborating uh together uh in the next series of podcasts. Thank you.

SPEAKER_03 54:12

Thank you for listening to our podcast from Startup Big brought to you by Tai Seattle. Assisting in production today are Isha Jain and Mini Varma. Please subscribe to our podcast and rate our podcast wherever you listen to them. Hope you enjoyed it.