Full Transcript

Dhruv Sharma: Listeners and friends, it's time to end Friday on a high. Today we're covering voice, we're covering water, water treatment, we're covering a couple of other things. There's a lot of news for us to break down with you. But before we get into any of that, I saw a really cool post from my friend Utsav earlier today. I want to ask him what that was all about.

Utsav Somani: He's going to embarrass me in the air today, Dhruv, don't make me do this. But I mean, my book is coming out. It's coming out on April 30th. It's out for pre-orders on Amazon. It's basically a compilation of these hundred interviews we've done of offline members. And I mean, basically sitting across these hundred founders who built these awesome companies, I think it's just amazing that the kind of insights that you can take and it'll be super helpful to founders across their journeys in early stage and mid stages to learn from these tidbits, anecdotes, insights and tips and tricks. And also the stuff that's not worked for them. I think that's what the key insights have been captured in the book. And I think it's going to be an exciting read for founders across all stages.

Dhruv Sharma: This is the first of its kind of book itself with like 100 founders sharing their lived experience, first hand knowledge.

Utsav Somani: All right. So let's actually get down to the real business. Bigger numbers being discussed by our big labs. Anthropic is at 30 billion ARR, up from 9 billion at 2025, 3x in four months and OpenAI, which is at 25 billion ARR, is first time losing its lead to Anthropic. And Anthropic has got 1000 plus business customers paying them a million each. This is insane. And Claude Code alone generates 2.5 billion for them.

Dhruv Sharma: Crazy revenue ramp itself. I think people tend to forget that they started 2024 at 100 million dollar ARR, 2025 at a billion. And we're just like a few months into 2026, they're at 30 billion already.

Utsav Somani: That is unimaginable. But let's talk about the new model as well, Mythos. Pretty crazy. Like, I mean, the previous one, which is literally just two months old, Opus 4.6, had near zero percent exploit success. And this one has 72 percent. It has literally found bugs in, I mean, a decade old software, right? And that is insane.

Dhruv Sharma: There's a news item today. I believe Scott Besant and the Treasury Secretary in the U.S. and the Fed chair, Jerome Powell, have called a meeting of like all the Wall Street CEOs to discuss the impact of this model. You think this is the AGI model or do you think this is still a pre-AGI model?

Utsav Somani: Could be, man. If they're breaking out of sandbox with this, I think it's scary, like it is absolutely scary.

Utsav Somani: All right, let's talk about the other big lab, OpenAI. They had a New Yorker investigation into them where a lot of their ex-employees have made allegations against Sam Ortman. They've recently taken over a media property as well, TPPF, which we were of course inspired by. And they paid a good amount for that. So I think they're making a big push in media.

Utsav Somani: All right. Let's welcome our first guest for the show today. Ganesh. Hey, Ganesh. Hey, how are you? What's up? Good to have you on the show. Thank you again for giving us the time.

Ganesh Gopalan (Co-Founder & CEO of Gnani.ai): Yeah, looking forward to it.

Utsav Somani: So let's start with the basics. I mean, you did start your company, Gnani, in 2016 before the full AI wave. Like nobody had heard of ChatGPT. Nobody had heard of Sovereign AI. We were not even thinking about AI at that point of time, right? And you decided to make a bet on voice being the dominant mode. So let's introduce Gnani to the world and give us that thought process on how this has evolved.

Ganesh Gopalan (Co-Founder & CEO of Gnani.ai): So like you said, when we started, AI wasn't fashionable. I think deep tech was a bad word. So for us, we were a couple of guys who are colleagues at Texas Instruments. And we said, look, we want to build out a deep tech company out of India. Rather naive when we started. I would say we started it in the most stupid way possible. We said we'll first build the tech and then the customers will come. I wouldn't advise any new startup founders to go that way. But that's what we did. We built India's first speech-to-text system. We bet that voice is going to be the best channel for communication between humans and machines because that's the de facto channel when it comes to humans, right? And very quickly, we built out India's first speech-to-text systems for many languages. Today, we have the largest company in this whole voice AI space, customers in India, various other countries, 200-plus customers. And generating what I'm happy about is a lot of superior business outcomes.

Dhruv Sharma: I'm wondering what was the first technical breakthrough you guys achieved in voice Ganesh?

Ganesh Gopalan (Co-Founder & CEO of Gnani.ai): I think the first one is always about data. And we started off with, we're from Bangalore, and we built the world's first speech-to-text system for Kannada when we started. And the kind of crazy ways that we got the data that we wanted to build this out was really interesting. So we came up — I can just tell you one example. So the IPL was just starting to get hot. And we said, look, there's a lot of bidding happening on this term, and people are looking for IPL. So we developed an app, we call the Indian Pronunciation League, which is called IPL. And that helped us as well with collecting some of the data that we wanted to. And I think the first time we realized that we are actually onto something and not just fooling around with some tech was when Samsung evaluated us. And then they said, look, you guys have better patents than we're seeing with international companies in this space.

Utsav Somani: And you went from speech to text to speech, and you removed the text element in the middle. So that removes latency as well for you, right?

Ganesh Gopalan (Co-Founder & CEO of Gnani.ai): Yeah, so that's a new model that we're coming out with. So we announced this in the NVIDIA GTC last year in San Jose. And essentially, what we're trying to do here is, we're building something called S2S. Essentially, today's traditional systems have, when you do voice AI agent, you convert your speech to text, right? Then you have some SLM or an LLM on the text, and then you convert it back to speech, right? But this has some inherent issues. So for example, when I'm speaking to you, you're not only watching the way I speak out certain words, the emotions, the pitch and the tone, but also what I say, right? But today's systems only look at what I say, because you're converting speech to text, but the way I say it is missed out.

Ganesh Gopalan (Co-Founder & CEO of Gnani.ai): I think the key thing is we've always believed that you need to own each and every element of your AI pipeline. So no API calls for anything. In our system, whether it's your speech-to-text, we do SLMs, we don't do LLMs, and then we do our text-to-speech systems, adding on to our own denoiser, our own turn taking LLM. I think when you're dealing with, especially for us, because we're dealing with global South countries, like India, the price points tend to be lower, I think you need to be very honest with your technology and build everything on your own. And that's the fundamental secret of success.

Utsav Somani: That is valuable advice to end the session with. Thank you so much for coming on the show, Ganesh. All right, listeners, we're moving on to the next segment. We've got Amrit of Indra Water with us on the show. Amrit, welcome to the show.

Amrit Om Nayak (Co-Founder & CEO of Indra Water): Hi. What's up? Hi, Dhruv. It's a real pleasure to be on the show.

Utsav Somani: Can you explain what the business does for our listeners?

Amrit Om Nayak (Co-Founder & CEO of Indra Water): So Indra is in the business of providing decentralized treatment for wastewater. And we're really different from others because we have moved away from the usual biological or chemical approaches, which are not really kind on the environment. They take a lot of time to treat water. And more often than not, they are not working. In fact, over 90% of the treatment assets in the country are down for over 40% of the time. So Indra is essentially targeting this gap where we have these electrically driven systems, which are able to treat water in a few seconds and give back most of the water at the point of generation for reuse.

Dhruv Sharma: Let's ask you to draw a distinction, Amrit. What's centralized vs decentralized treatment?

Amrit Om Nayak (Co-Founder & CEO of Indra Water): Centralized plants rely on the principle that the wastewater from every building, every establishment, every industry needs to be collected, needs to be transported, treated and then sent back for use. So there are different aspects to this value chain where a lot of resource, a lot of time, a lot of infrastructure needs to be built. Decentralization looks at this problem statement a little differently. What we believe is that if we could have systems which are truly modular, which are really small and compact and could fit into these establishments without requiring a lot of changes, we could actually use water effectively at the point of generation. So you're eliminating these massive pipelines, pumping stations, a requirement for a large tanker network or redistribution systems.

Utsav Somani: And that's why you've been able to win over blue chip plants, right? You've got Unilever, you've got Aditya Birla, you've got Tata as well.

Amrit Om Nayak (Co-Founder & CEO of Indra Water): I think all of them are big ones. Unilever was the fastest adopter among the lot because of the way their entire business functions. They have localized manufacturing to deal with market demand. So most of these units are distributed, not too large. When it comes to Tata, we've been working a lot with Tata Steel. That's where we started. And steel is a massive water guzzler. You could be using up to 3 crore or even more liters of water per day in a single facility.

Utsav Somani: As a final closing question, you've raised 4.28 million in this space, and you've paid 12 crores in revenue. What does the path to 200 crore look like for you?

Amrit Om Nayak (Co-Founder & CEO of Indra Water): In fact, yeah, we did raise money in Jan 2024. That was our series A. And we are closing 15 plus this year. The upcoming year, we're looking at 50 plus and then getting closer to 100 the year after. Because we're in a position where we can deliver and we can control a lot more of the value chain than we could do in the past.

Utsav Somani: All right, wishing you all the best on the journey. Thank you so much for coming on our show. All right, our final guest segment, we have two names today joining us. We've got Devaj and Avik as well. So welcome to the show.

Devarsh Saraf (Founder & CEO, Bombay Founders' Club): Two years ago, I moved from law to building in legal tech. So I moved from being a lawyer to being an entrepreneur. I searched for the best communities in India and I could not get into Offline. So I started something called Bombay Founders Club.

Utsav Somani: I think Avik is here. Avik, let's welcome you on the show. You took a call in the middle of a live stream. Like, is that how you do things now?

Avik Ashar (Principal at Artha Venture Fund): Unfortunately, when there's certain people who call, those calls have to be picked up.

Devarsh Saraf (Founder & CEO, Bombay Founders' Club): Now we've gone to a largely micro model where we're focusing on getting a few relevant people in a room together in the most frictionless way possible. So the entire goal of BFC is to help founders build their social capital as frictionlessly as possible and with as little effort as possible.

Utsav Somani: Avik, I mean, I want to ask you something. You've scaled up Zillingo across five countries. You've invested in over 50 plus companies across Riverwalk and Artha Ventures as well. So how do you view community businesses from an investor lens?

Avik Ashar (Principal at Artha Venture Fund): Community businesses are something that are integral. You cannot look at them purely from the lens of, hey, is this company going to give me that 100X return? I think that's the wrong way to view these businesses. It's, and I'm a big fan when I'm investing of investing in a healthy mix of companies that have a binary risk, which is companies that can give you that crazy return versus companies that are going to become part of the infrastructure, become part of your social fabric.

Utsav Somani: And you've written that VCs are founders too, and most people don't understand the economics of fund management. You're pretty vocal on LinkedIn, and you also said that you don't do diligence on investors. So care to elaborate?

Avik Ashar (Principal at Artha Venture Fund): Without obviously naming anybody, you cannot take capital and cry about it later. A VC founder relationship lasts longer than the average US marriage because once you've got somebody on your cap table, they're there. So it's crazy that you are willing to sign up for what's probably a 10 year plus relationship, and you're doing less diligence than what you do on what to order for dinner.

Devarsh Saraf (Founder & CEO, Bombay Founders' Club): I'd say as a community builder, you shouldn't ideally look at your members as customers. I think that makes the relationship transactional. We're not here to build something transactional. I often call myself a glorified phone operator. The magic is the members, product is the members. The only job of a community builder really is A, to ensure that the curation is on point. The day they look at themselves as customers, you're finished. They have to have a sense of belonging with the community.

Avik Ashar (Principal at Artha Venture Fund): I remember about three and a half, four years back, they said, hey, we're going to remove the Indus entrepreneurs and just squeeze it down to Thai. And I was asking, hey, why are we doing this? And they said, this was started to help South Asians, but that's not the goal anymore. Thai has moved to become an organization to help all entrepreneurs anywhere. And the way I've seen Thai create women entrepreneur programs, create programs across the world, that kind of community is what I've loved.

Utsav Somani: Amazing. Here's to cheering on and finding your tribe. Thank you so much, Avik and Devaj for joining us. Have a wonderful weekend ahead. What's up, Dhruv. Thank you.

Utsav Somani: All right, listeners, today went off without any technical glitch. We've had four awesome guests on the show. We'll see you on Monday. Have a wonderful weekend. Stay safe. Bye-bye.