The Exchanges

Q Uh, and, and do you practice daily also?

A Yeah, yeah. Like, come what may, like morning, 45 minutes, To one hour is a given, right? So I, I don't do it twice a day and it's usually recommended you sit in the evening as well, but I'm super tired by the time I, uh, get to, uh, get to that. Uh, but yeah, uh, on a, in a year, at least two 50 days in a year, I would sit in the morning for an average of 45 to 50 minutes, right? So, uh, and you also asked how did, how does it help the company and stuff, right? So, uh, Uh, so surprisingly, a lot of my folks in the company came to me and said, you have changed.

AI assessment note: “Yeah, yeah. Like, come what may, like morning, 45 minutes, To one hour”

Q Then, then probably You know, it's a tangential thought that a roll-up, if you raise two hundred million in roll-ups in India, and acquire five to six companies in the US, you will be able to acquire more than two hundred million in, in, in revenue, then possibly take this structure public in India. Why has not happened?

A Yeah, so I think the value is, like I said, you will trade on EBITDA multiples, right? Like, look at us, we will eventually trade on EBITDA multiples. So if I were just putting these 50 companies together, and let's say they were all these four or five percent margin businesses or no percent margin businesses, You will not get the value, right? Now what we're doing is essentially you're saying, okay, this is an also run, uh, I mean, it was a company which was great at some point in time or whatever, right? Uh, but their tech stack hasn't evolved or something might be the thing there, or they're sitting in an entity which doesn't think of them to be strategic enough, whatever is the reason, right? So, uh, now the value in the capillary, and this is, uh, there in our DRHP as well. Comes from actually upgrading those customers from that platform to ours, where suddenly a 15, 20% gross margin business becomes a 60, 70% gross margin business, which then allows us to reinvest in the product. So the customer constantly keeps getting a best in class product. Unlike in a five percent margin business where the business is barely able to, ah, you know, what we've also done is a bunch of these companies that we have bought were very services heavy. Like, you know, to serve one customer, you would need 15 to 20 employees.

AI assessment note: “if they were all these four or five percent margin businesses... You will not get the value”

Q And, ah, one thing unique about the capillary culture, since we talked about that, and, ah, your team gets 10 day, ah, Vipassana leaves also, right, is out of seven 30 members in the team, hundred plus have stayed for more than 10 years. What has led to this kind of a culture?

A Yeah, so that's, I think I've been extremely lucky as a founder, uh, you know, to have, like, really, really good team members, uh, all along. We are now, 1718 years old, right? We're 17 years as a firm now. Uh, and out of the seven 30 employees, about a 130 have spent more than 10 years with us, right? So it's, uh, uh, like, if you compare that to Bangalore, you know, I think we are like five X better in terms of retention. Uh, our monthly churn is one percent. One percent of our team leaves every month, which is very low. Till COVID, it used to be five percent a year, right? So I think a few things that have worked for us is, Uh, you know, one, uh, we're not the best paymasters. You know, we are, like, there's always been this belief that when we started Capillary Off, a bunch of people who joined us were kind of people who knew us, not necessarily personally, but would have, you know, worked with us at other, you know, other companies that we worked or were juniors from IIT Kharagpur, Ah, and so there was always this feeling of a community more than a company.

AI assessment note: “there was always this feeling of a community more than a company.”

Q That's quite a lot of difference. And what are the companies that capillary alums have started?

A Yeah, we've had a ton of, ah, you know, I think we've had like, 50 plus companies, ah, I don't know, probably more actually. Like, very, very, ah, You know, I think it was 2012 or 20 11, when one of our very early employees went and started up. And we were extremely pissed. Both KK and I were very pissed. This is Shubh, Shubh actually. Shubh is one of the co-founders at NPL. And he was our, he was our second employee, right? So when he started off, we were like extremely, what's going on? So, and to our bad luck or to his luck, when he said he'll start, a couple of good folks also said we will also join him.

AI assessment note: “Shubh is one of the co-founders at NPL. And he was our, he was our second employee”

Q And, uh, what, what tool did you use to build it?

A I mean, I built it on, like, I, I, I've actually tried both now, right? So I built it on Gemini initially. Then I built it on Claude to see which one's better. It's on Xcode. So it basically uses the Apple, you know, visual framework and all that stuff. But, uh, and look, Uh, like, there is a real change, right? Like, people like me who last wrote code in MATLAB in 22,008, not even 2008, 2006, Are now like can build stuff, right? So it is a, uh, so the fear is real, right? Uh, now unfortunately what's happened is I think markets have painted everything in, in one paintbrush, right? In one flat paintbrush of saying all software is going to die. Uh, I don't think that will happen. So at least we have a, like the framework in my head is, uh, you know, are you a system of record? Are you a system of engagement or workflow? Or are you a system of intelligence? I think The system of records will still stay. Like AI or no AI, your bank ledger will still be a bank ledger. You still need a hundred percent auditable, hundred percent accurate, uh, platform, uh, to deliver your trade. Uh, so if you're not a system of record, start moving towards a system of record very quickly. Uh, because I think AI will replace workflows. You know, AI will replace, uh, all of that intelligence. Like we have some beautiful bots now, which can do the work of an analyst. Uh, like sitting in a, uh, agency ou…

AI assessment note: “I built it on Gemini initially. Then I built it on Claude”

Q You know, in your journey, if you have to attribute the, the current success of capillary to five major decisions, what would they be?

A Uh, five major decisions. What would they be? Uh, wow. That's a very deep question. Uh, yeah, first is, uh, uh, I think when we started, uh, we did some things very smart, right? So, uh, that whole period of, you know, uh, not just going to build what we wanted to build, but we spent six months speaking to customers before we wrote a line of code, which I think was a very, very good decision because by then we knew who our first five customers are. We weren't selling, we were listening, right? A lot of that played out very, very well, right? So the initial part, I would say that was a, Uh, that was a very, very big, uh, big aha for us. Uh, the second decision I think was, uh, you know, our, um, like, uh, our ability to show value, like, uh, quick value delivery, right? That in three months I can show you how you, uh, meant that Not only did our India customers roll very fast, like customers like a Puma, customers like them, actually took us international as well. It's very funny. Our international journey didn't start with us wanting to go international. Started with our customers saying, you know, what you solve for us in India, can you solve for me in Singapore and Dubai?

AI assessment note: “we spent six months speaking to customers before we wrote a line of code”