Full Transcript

Dhruv Sharma: Hey there, listeners. How is Monday coming along for you? This is stream 126. Today, Utsav and I are chatting with Suruchi Rao, who's the co-founder of Ossus Biorenewables, a Bengaluru-based clean-tech company. Suruchi, welcome to the offline network. You. It's great to have you with us. So please tell us what the company does.

Suruchi Rao (Ossus Biorenewables): Sure. To distill a lot of fairly serious engineering, let me put what we do, the way we would talk to a customer. Right? So let's say you're, quote unquote customer for Ossus Biorenewables. You've come to us saying, guys, the requirement for green hydrogen, it's difficult for us to source it from the market today because it's too expensive. We've heard you can help us use our own wastewater stream to generate green hydrogen. What do you think about it? And please say yes. What kind of wastewater do you have? If it passes certain checks and balances, please say congratulations. We can make you, a lot of green hydrogen that's priced approximately one-sixth of what you would pay out there in the market. And the best part is that the supply chain involved with procuring this green hydrogen fits with you. It's internal because the feedstock or the starting material for the green hydrogen is your own wastewater. So to boil it all down, we're a green hydrogen company, but unlike most people out there, we use feedstock that sits at the customer's own site, which is his own waste to the cost center, and we convert it to something of value.

Utsav Somani: And what is green hydrogen and what is it used for?

Suruchi Rao (Ossus Biorenewables): Oh, good question. Let's start with green aside for a quick second. Right? Hydrogen is a gaseous molecule. It's extremely light, which means a kilogram of hydrogen probably occupies the, you know, twenty twenty odd thousand square feet space. Half of it would be, you know, covered with hydrogen gas because it's extremely light. But while it is difficult to handle because it is extremely light, it has a lot of energy packed into it. Basically, what keeps the sun going is that dense in energy. And when you want to transition from using fossil fuel, especially in sectors that require extremely high energy requirement for heating, heat generation like steel, cement, chemicals, you have no option but to go with hydrogen gas because the green energy coming from solar or wind will never provide the energy density that hydrogen provide. Now that's the primary purpose with which the consumption of the demand for hydrogen has grown as we push for decarbonization. So So if you were to look at hydrogen itself, it's nothing but an energy dense molecule. The term green comes in when you produce hydrogen without generating any carbon dioxide. That's basically how green energy works in general. You want to be able to solve for what fossil fuel does, which is to add greenhouse gases to the environment, makes life difficult for all of us and particularly difficult for the future generation. Hydrogen burns without generating any CO2, but the production of hydrogen itself does involve CO2 if you were to use fossil fuels. Green hydrogen is when you produce hydrogen without any carbon dioxide.

Dhruv Sharma: So, Suruchi, traditionally or conventionally, how have they produced, green hydrogen? I'm I'm curious to know how are you able to do it at one six of the price with your bioelectromechanical process.

Suruchi Rao (Ossus Biorenewables): Sir, the way people talk about green hydrogen, the first patent for green hydrogen as we know point, dates to the 1900. And it's one of those 10,000 leagues under the sea sort of a dream where people talk about basically breaking down water to generate hydrogen and oxygen, H2O. It's broken down. The problem with this is that we all know how stable water is as a molecule. You can imagine how much energy it would require to take liquid and convert it to two separate gases, basically going back all the way to how the planet started. So in terms of energy consumption, you're talking about using anywhere between 50 to 55 units of energy. Now let me break that down into something similar. For those of us who probably have dishwashers, multiple, you know, washing machine dryers, our energy consumption or the electricity bill that comes to us is, you know, on the higher end. It's 200 units. That would produce you four kilograms of hydrogen. Four kilograms only. That doesn't really yield much in terms of deals a lot of energy, but not enough at an industrial scale. So green hydrogen when produced by splitting water requires a lot of energy. If that energy is not available free, quote unquote, because it comes from renewable energy like wind and solar, you're basically talking about being addicted to a system where you're convinced or you have blinkers on saying water splitting is the only way I can make green hydrogen. It doesn't matter to me that the amount of energy required to produce a kg of hydrogen is too much. At the end of the day, industries can't afford to put those blinkers on. If 200 units of energy is only gonna give you four kg and their, you know, average hourly requirement is 10 tons, and I'm saying a very small number. That's 10,000 kilograms of hydrogen that is required on a hourly basis. How much energy that is, you can do the math yourself. But that's the way people look at green hydrogen today. The way we are able to make this entire process much much one-sixth the price, Basically, because we don't split water. We do use electrons and protons combining to make hydrogen gas, so we still rely or lean on electrochemistry. The source of our electrons happen to be biological. We use biological catalyst to be able to break down the organic content in industrial wastewater and generate hydrogen gas. The energy requirement in our case is top line 3.8 to four units per kg. We're talking about the one-twentieth energy demand, and the feedstock itself does not have to be sourced from the environment. You don't have to buy dirty water, make it clean, split it into hydrogen, or you don't oh my god. Imagine having to buy groundwater to make hydrogen gas. We don't have to do any of that. Waste sits and exists already, which is using a small amount of energy, well selected group of microorganisms, and general electrochemistry to be able to produce hydrogen at one-sixth market.

Utsav Somani: Fascinating. I think, so before we go into more specifics, I have an energy related question. We had the founder of Solar Square, Mars Stream as well, and, she explained the solar industry very perfectly to us. So I think on a spectrum, like, suppose solar is cheaper to produce per unit of electricity

Suruchi Rao (Ossus Biorenewables): Mhmm. And,

Utsav Somani: nuclear probably, I'm guessing it's is that the other end? If there's technology for it or if it makes financial sense, like, what does that spectrum look like and where does hydrogen lie in that spectrum?

Suruchi Rao (Ossus Biorenewables): So interestingly, if you were to look at any IEA sort of projections for what is the distribution of energy. Right? You will see something like solar, wind. You can have biomass there, and biomass basically indicates agricultural waste. So trees that you grow also hold energy from when they consume the CO2 in the environment. So you'll see biomass feature there, and then you'll see nuclear, and then you'll see other. Hydrogen usually sits there because people don't really look at hydrogen as a source of energy. Now let me break that down for you. When we talk about industrial energy use, most often, you're talking about generating heat or using hydrogen as a raw material. The power that you and I talk about, it's the kilowatt hour number. If you notice, hydrogen itself is produced by combining electrons and protons. So there's electricity involved in the making of hydrogen. And then when you convert it back into electricity, you need some sort of engine to convert it back into electricity. That happens to be something known as fuel cell. And fuel cells have efficiency which are poorer than IC engines. You're talking about a 30%, 25% conversion. And so you spend all this effort or whether it is with our hydrogen or with splitting water. You've used 55 units of energy. You've converted into hydrogen. Theoretically, you're supposed to get 35 units of energy, and you end up getting 14 at best at best. So there's a lot of energy which is lost in the entire system. That is why in projections when it comes to, how the energy landscape changes, if decarbonization was supposed to be adopted, you know, at a wide scale, you will always find hydrogen in the other's category. So it's not that it's difficult for me. You always will end up thinking about energy, from a input output perspective. Right? Because that's where the money is. If I'm spending more to make something, I might not usually use it as the best way to go about whether it is decarbonization or whether it is a simple enough energy source. It's difficult for people to lean into that. So for this to basically make that jump, you need solutions like ours that are able to produce hydrogen at one-sixth the cost. So you have to change the status quo. You need to be able to see the impact of hydrogen. But today, I would say it doesn't even sit down the spectrum. It's a molecule of the future. Unfortunately, it's been a molecule of the future since the 1900. So disruption is much needed in this space.

Dhruv Sharma: So, actually, we're looking at your background and it it appears as though those those look like high precision systems, and, have you guys built them, in house? Are there have you imported them? And if it's the former, I think a lot of founders increasingly getting interested in hard science and engineering. How do you build stuff like that when the domestic sort of component ecosystem is still maturing?

Suruchi Rao (Ossus Biorenewables): Very interesting question. So, you know, this is this is something that we always tell ourselves and tell our investors to, for a company that leans heavily on knowing what the process is, basically, we've created a new process. Right? The this way to make our decision didn't exist before. We are not a process company. We're a product company. So you will find that a majority of the engineers, so we're a team of about 50 people, 37 of them, you would say fall under the hard engineering, mechanical engineering category. Because as you see behind me, we've had to make a lot of these things ourselves. When people talk about not having access to equipment, and most often people talk about the countries somewhere, engineering components have to be imported, and there's an entire discussion about that. We because we built fundamental science, we didn't have vendors out there who would do what we want them to do. Layer that on with the need to move faster. So we have deployed our solutions from 2021, three years after setting up the basic science for this. We were on-site with Tata Steel and Jamshedpur, and we didn't really stop, which meant that even if now it feels like we had to reinvent the wheel in some of these cases, we do have to do the hard engineering, which is so it was a lot of nuts and bolts and being on-site for ninety days in very tough conditions. And I think that's what sort of set us apart from most people out there, in the green hydrogen space where when they talk about the levelized cost of hydrogen, they're talking about, I'm going to do x y z to reduce my dependence on membranes from Germany, compressors from China, walls from x y z. We haven't really gone about it that way. We'll see our vendor list, which is extremely hyperlocal to Bangalore. And this is something that's interesting. Right? We have customers who sit in Gujarat, and we see the kind of chemical engineering support or infrastructure that is available. But we did something that was different. We combined biology, electrochemistry, which meant that we couldn't really go out and see customer or have a large vendor list. Our bill of materials is completely internal to us, and many people don't realize Biocon sits out of Bangalore. We're a biology company. Biotech is the heart of, Bangalore, and that's the Bangalore that we know. So it's quite different in our case. We've never really had challenges with, oh, there's import dependence in this case. You will not find that with us. What you will find with us is how do we go about vertically integrating this because this means that we have now started to become an engineering company, which resembles, let us say, Rolls Royce back in the 1800. When you're the first one doing it, you'll have to do everything yourself, which is why you see a lot lot of the basic science and the engineering going into our solution. What goes on the ground looks beautiful, but a lot of hard work that goes into it, and we have to thank sort of ecosystem that exists in Bangalore. Many people don't know that they're quite

Utsav Somani: sure. And you mentioned the Jamshedpur, steel plant that you're working at. Did they have, like, a minimum viable influent stream that you look for, or are certain industries more, or their factories and their industrial setups more geared up for, like, being suitable for your offering? Or is it open?

Suruchi Rao (Ossus Biorenewables): Interesting. Very interesting question. And we've we've decided to sort of when we started, we decided to take a more, what is my total addressable market and what is my serviceable obtainable market. Right? So for us, the the edges of this entire thing when it comes to building the platform is actually the market size. So if you were to if if you were to ask us, we basically divide our market into two categories, and each one represents a different wastewater stream. So on a platform level, you're talking about taking in organic matter, converting it into electricity, and then to hydrogen. You can do that with any sort of waste. You can do that with waste water. You can do that with agricultural waste or biomass. You can do that with, compressed biogas, unit waste. So what is it how is it that we how do you choose? Right? In our case, we went with two kinds of customer. Once that we know have decarbonization as an internal goal. So these are the guys who have budget set aside, and they know I have to produce this much hydrogen at such a benchmark price by 2030. We do understand we don't have time until 2030 for them to get mature and start consuming. But we do know that they are the ones that we need to invest in relationship building for the long term. So we have to demonstrate our solution, and that's where Tata Steel did. Their decarbonization timeline is 2070. And how do you break that down into five year terms and build your solution with them is basically what we've done. In terms of our at at the risk of using Hindi here, customers, you know, the guys who actually pay us on a monthly recurring basis, very different sort of wastewater. These are the guys who use hydrogen as a raw material until today. So they're not using hydrogen as a decarbonization molecule. They have a current requirement because unlike what most people believe it's there are people who use hydrogen today. They have to. It's a chemical molecule, and they dip into the local market where they buy cylinders. Coincidentally, their wastewater happens to be extremely organic rich. Now the reason why both of these would fit into the wastewater profile for us is that wherever there is hydrogen consumption, know that there will be organics that are bleeding into these wastewater systems. It's at the end of the day, the process industry. You take something, convert it to something of value. So whatever is the waste or the loss, it is what goes into the wastewater. The only thing that distinguishes this is the guys who use hydrogen today and who pay us the recurring revenue, the roti capram account, that's a serviceable obtainable market. The TAM is the decarbonization value. That's the 90,000,000 ton target that IEA has put down. Nobody can find the road to it. We're slowly chugging along, sort of nipping away at it five years at the time.

Utsav Somani: I mean, you don't identify yourself as a climate company in that regard.

Suruchi Rao (Ossus Biorenewables): In that sense, we identify ourselves as people who make hydrogen accessible. It is a little jaded for us to say this. ESG, driven business is is on the down and down. We're still a company that has to go out and raise funds. But at the end of the day, even if you were to set the decarbonization or the ESG goals aside for two minutes, we are a technology company. Right? We built a platform, and we need to get that validated in the market as quickly as possible. Otherwise, you'll constantly remain in the R&D loop of being in the lab to get your, sort of, badges you need to go down into the field. And if it's the guys who want to use hydrogen today, imagine a solution which drives decarbonization without actually naming it, and that's us. So all the guys who switched over to us have switched over to green without without having to worry about the price tag or having internal ESD goals in the first place.

Dhruv Sharma: So you already have very supportive investors, but let's do this thing that sometimes called reverse pitching. Let's say the next time you're out in the market trying to raise around, what would you expect from your investors? I mean, even the prime minister made a reference to the bioeconomy and how important is, in his Independence Day address. So, yeah, tell us what you'd like in your investors the next time you're raising around.

Suruchi Rao (Ossus Biorenewables): Oh, I don't know if I can be that candid given the fact that we're in the middle of a fundraise at the moment, but I think patience is extremely key, especially when you're building in a space which is uncharted territory. Right? I know we've been very lucky with Gruhas and Rainmatter, and they've showed us that patience. But now we have to put the big boy pants on to talk about every time we talk about multiple crows, it it gives us heavies, geebies. But even as a solution is moving from lab and it has some primary validation, then it gets to some kind of demonstration scale where it's somewhere commercial. Let us see here at a technology, the readiness level between seven and nine. Whatever comes next is just us having the discipline necessary to repeat what we've done. But there's still some patience, you know, that is required because you're talking about a market that is shifting on a daily basis. We've seen how how little decarbonization is able to earn. We got lucky in that we never started that way, and therefore, we're able to segue rather quickly or spread it our way around the pace. But patience is something that we would definitely expect from our investors even if it is. Patience is different when you are at an early stage. It means you're allowed to make as many mistakes and still find as long as you find the light at the end of the tunnel. Patience at a large scale is supposed that invoice that was supposed to be due on the fifteenth of the month gets pushed to, you know, fifteen days later. When revenue is king, we're learning how to do this. And as first time founders, I would definitely say patience from the investors to be able to find, your footing in a new market that is being created would be fantastic. But, otherwise, I think the Indian ecosystem is beautifully poised. I don't think they've had any concerns about fundraising, funding mentor, or found us as in a situation where we had to justify why we exist. No one has once asked us. No one wants to decarbonize anymore, guys. What are you gonna do? So that should that's good.

Utsav Somani: No. No. Wishing you all the best. But I think, I mean, let's quantify all of this. You said the industry is shifting daily, or there are changes happening daily. Like, is it in terms of unit economics? Is this, I mean, some technological change? Like, what is changing, and how does the business look right now? Any number that, of course, don't have to disclose anything that you're uncomfortable with?

Suruchi Rao (Ossus Biorenewables): No. Let's let's talk about something that has moved, that has proven extremely valuable to us and something that has not. CBAM or just to break it down, the tax that is now going to be imposed on goods that are being imported into Europe is a dollar price that is added to every ton of aluminum or steel that leaves India. And that number is directly related to the emissions that come per ton of producing whatever was that metal that was exported from India and imported into Europe. Now that, there is a the bureau value assigned to this, it's worked really well in our favor because people don't look at hydrogen as hydrogen alone. You get to add, CBAM credit. I will we are not very in favor of carbon credits because it doesn't, earn much as value, and then there's this question of where do you trade it, who owns it, etcetera. But CBAM is something that we are helping our customers sort of leap over and say that, look, even if you have to blend blend 20% hydrogen into your natural gas stream. So we're not asking you to decarbonize overnight. Right? Do this slowly and see what impact it has. So while I can't mention numbers because it depends on the company and where they trade, it's been it's been great to sort of, when you're putting the unit economics together, it's great to sort of leverage that. And levelized cost of hydrogen now includes that as an additional line item, which is fantastic. What has not worked very well in our favor is that I'm sure you've seen what happened with the last call. Right? So the more at a national level people are willing to kick the can down the road the effects of climate change are I'm I'm sweating. I'm in Bangalore. Right? Effects are there. The people who need to be making decisions to do something about it want to prioritize common man's life. I get it. We get it as a company, but it's not really helped when that next contract that we've designed with another steel company. Say, the third one that we want to bring on board is, like, guys, it's not something we need to be concerned with at the moment. Let's work on wastewater treatment as a solution. It's something that we do as part of hydrogen production. We're a hydrogen company. And you can see us, you know, starting to think about, should we use on a wastewater treatment company? Seems like it's gonna earn us more. We shouldn't have to be making those decision. But when I say the market is shifting, this is a political outcome and is likely to change with time and effort. When I say sort of the market is for us, we feel it on a daily level. So we're an eight year old company. And at Tata Steel alone, we've seen, you know, four managers change. We have ONGC as a customer. You have people that get transferred every five years. So for any startup out there, the things who have signed this customer on next thirty years, I'm I'm set. You know? I've spent this is this is my cash. My lifetime value is 30. Unfortunately, that's not true because decision making is something that changes on a day to day level, and this is a problem that happens with B2B companies. To be aware of it beforehand, I think would definitely have, pushed us to closing something sooner. Now, obviously, we've burnt our fingers. And your champions may not necessarily stay forever. So relationship building in B2B, I'm not saying anything new, of course, at this time. I'm just speaking as though as though I'm the first person who discovered it. But, when I say things shift, urgency changes with the person who, joined on to take the conversation forward.

Utsav Somani: Dhruv, any final closing question?

Dhruv Sharma: Yeah. I guess one on circularity, Suruchi, which is, like, when you talk about Tata Steel, for instance,

Suruchi Rao (Ossus Biorenewables): instance,

Dhruv Sharma: you know, and and Jamshedpur, it's abundantly clear how, the effluents themselves are being used as input and, you know, how everything is playing out. But beyond these carbon intensive, heavy emission, heavy industries, where else might green hydrogen find application? I'm I'm also thinking about the agricultural industry, right, which can

Suruchi Rao (Ossus Biorenewables): Mhmm.

Dhruv Sharma: Which will feed the most amount of feed stock presumably. Is there any

Suruchi Rao (Ossus Biorenewables): Mhmm.

Dhruv Sharma: Outcome that they can hope to expect from solutions like yours over the next few years?

Suruchi Rao (Ossus Biorenewables): Absolutely. The things that we don't don't notice and that Europe has done very well is that it has not niche applications for hydrogen in moving goods. Right? Hydrogen is really good when it comes to EVs need frequent charging, it gets discharged. Hydrogen has the capacity to just keep going on without meaning too much in terms of the weight per unit of movement of material. Heavy lifting, warehousing, a lot of these things come into play if you were to look at agriculture as an ecosystem that not only produces food or good value, but also its waste represents a source of energy. But all of this basically involves movement by vehicles that can always be powered by hydrogen. We just don't have exposure to systems that make that happen. So there's, again, a huge gap where mobility, which doesn't look at, mobility from moving you and I across Koramangala to Indiranagar, but mobility when it comes to moving goods is something that hydrogen can do really well. If there is sufficient demonstration that there is hydrogen available that makes it cost effective, I think that transition is, but a matter of time. But this is just one small thing. You can do a lot of things with hydrogen region when it comes to long distance mobility. You can talk about stationary storage. I know a lot of my friends in this space are working in this area, and I know that's gonna become the next big thing. But industrial use case is very large by volume. But this is where I think the growth will happen slow and steady.

Dhruv Sharma: I think they've also been trying experiments in aviation turbines with, with, what they call SAF, sustainable aviation fuel. Is is that also an example of

Suruchi Rao (Ossus Biorenewables): Absolutely. So you need hydrogen to be able to make SAF, and that's one of the key areas identified by the government as well as where will you use hydrogen? SAF is one of the places where hydrogen will find imminent use. It needs three to four players to come together and form a consortium enough for this to take off. But that's, like I said, a matter of time.

Utsav Somani: There were some auto automobile makers are thing also doing tests on hydrogen as a fuel. Right?

Suruchi Rao (Ossus Biorenewables): Yeah. Just this weekend, I think, also, I did something that was a 100% hydrogen powered turbine. So, yeah, good timing here, I guess.

Utsav Somani: Alright, Suruchi. Thank you so much for coming on our show. Wishing you and the team a very, very best ahead.

Suruchi Rao (Ossus Biorenewables): Thank you. Thank you so much. You had a lovely time.

Utsav Somani: Thank you.

Dhruv Sharma: Thank you.

Utsav Somani: Alright. Listeners, I think we're waiting for our second guest. I believe we have some new cool projects to showcase. There was a whole bunch of AI driven projects. I think people are hacking, some awesome stuff together. We saw this trend, kick start with this, salon Deluxe Salon, music player, which was playing, like, music from nineties. I mean, it's got beautiful UX, UI, and you can just, queue up music and stuff and, huge bunch of nostalgia built into this whole thing. So it did go viral. But what it triggered is, I think, a lot of people are now building some exciting cool projects. Did you discover this, Dhruv? It must have popped up in your feed.

Dhruv Sharma: It popped up. It reminded me of my hairdresser, Banwari, used to give me 15 rupee haircuts, and all of us have stories like that.

Utsav Somani: I mean, beautiful design. Right?

Dhruv Sharma: Yeah.

Utsav Somani: Beautiful design. And 32 people are online actually listening to Kumar Sanu and Alka Yagnik right now, it seems. Then this one popped up as an independent state project. It's called India isbuilding.life, and, it's quite fascinating. What you can do is, choose I mean, exciting things in energy, infrastructure, transportation, public systems, agriculture, and a bunch of other things that have happened in the last forty eight hours or since, whenever they've been measuring. And you can pick an industry, like, say, infrastructure, and you can see what's happening. Like, I mean, are people responding signals? Like, what's happening in the bullet train? I mean, it's, I mean, nice. Like, I think it's just a good way for India to sort of visually capture all of this and for, readers to be aware of what the country is doing towards, their goal of 2047 Viksit Bharat. Very cool project.

Dhruv Sharma: I think we'll make a similar map for, TON appearances as well. Also, I'm very curious to see.

Utsav Somani: And that actually can be nice. Yashree, please take notes. Pranav, please take notes. We might actually do a map of where all the guests are located, at least the series, whatever we can pull from the LinkedIn. That can be quite cool. Show up their bios here and stuff. And this is another one that, popped up. Apparently, six twenty three reports of drives across the country. And you can do a search across department, cities, compare, and I don't know, know before you do. So I don't know what the state I mean, statistics look like, but it's just a way for citizens to hold their, country account accountable and also the ministers accountable or at least report, where they face trouble. Yeah. I mean, interesting that people are building all of this in one single weekend, and, yeah, I mean, this is good use case of AI and citizen journalism, to some extent or citizen. What would you call this as?

Dhruv Sharma: I would call this a reinventing general apart from first principles, but that's on a on a on a lighter note. I think this is incredible. In some of the comments, they were quite funny as well, but, hats off to whoever made this. Really cool.

Utsav Somani: Apparently, it's going around, WhatsApp groups as well, so kudos. It takes Harish to put something like that out. And then, of course, is a map of Bangalore, basically. You can see literally where all the startups are, where which we see is funding, this thing. Apparently, the creator, Aditi, got so many things that they had to add, I think, 200 or 500 startups over the weekend itself. Now they started selling ad spots as well. So you can buy an ad spot on this for, I think, 2,500 for twenty four hour flash slot.

Dhruv Sharma: Yeah.

Utsav Somani: And pretty cool. Like, I mean, it's nice as a side hobby, and I think we would encourage all our listeners, listeners, I think, to do this. Right? Through like, just pack together something. Like, I think it gives you a real taste of what it takes to put something out there. Of course, this is all the software stuff, but let us know if you like this segment where we actually showcase two, three interesting things in a week and call it, like, a TON showcase where you get to see and hear from us about these interesting projects. And then, you'd like us to cover hardware, you'd like us to cover a specific segment or sectors or industries, we'll definitely bring this up, as well on the show. And couple of big acquisitions closed on the weekend as well. One is reported. One, of course, we had covered earlier. SpaceX is doing, Cursor. They've acquired and closed that purchase for 60,000,000,000, and Stripe is doing open router. It's still a reported, deal. Bloomberg reported on the weekend for 7,000,000,000. Break it down for us.

Dhruv Sharma: The open router one? Yeah. Yeah. I I think it's a, you know, about two years ago, everyone thought that the the model universe will have very high fidelity. In other words, every, individual and every application developer will only work with just one model, but all of that changed very, very quickly. And so what OpenRouter built essentially was an AI gateway that sat between applications and model providers. And, it was it was it was a unified API that gave you access to hundreds of models with a single endpoint.

Utsav Somani: 400. Yeah.

Dhruv Sharma: 400 and and, you know, they keep the numbers keep growing. But it wasn't just like an ordinary unified API. It also had, like, model of their behavior. So for instance, it was doing model selection for you. It was doing token metering and eventually, you know, you'd have to pay up. And so even billing and, of course, before that, auth and and so and guess what? Who was as OpenRouter was doing the token metering, who was taking care of billing? It was Stripe. So Stripe was

Utsav Somani: billions to their transaction volume, I guess.

Dhruv Sharma: And their stated mission has always been to grow the GDP of the Internet economy, and now the Internet economy is the AI economy. And, of course, they will have a role to play in it, and so that's why this, you know, reported news I was actually going through Stripe's 2025 annual letter to see if the brothers had put down any hint of this happening. They had their great they had their own point of view in agentic commerce, but if that's the future of the Internet, right, where, on the application side, agents are gonna pick which models to, to use, you know, coding agents, for instance. So, it only makes sense for for them to do the SACWIS. The other interesting thing is lots of other companies have their own AI gateways now as well, including Cloudflare and Vercel, and there's at least four or five others.

Utsav Somani: I think some of the other interesting things that stood out to me personally. So it's basically a high volume business and a low margin, game. They're taking 5% commission, which accounts for 50,000,000 in revenue right now in annualized revenue. So Stripe is adding billions, like, much like their main core business, which is payments processing. It's, like, extremely high volume and low margin business. But, I mean, there are a few other things. Like, I mean, a company valued at $159,000,000,000 is acquiring a company, which was actually valued at 1,300,000,000, previously, and now they're, bumping up the price in a very quick, duration to 7,000,000,000 as the acquisition price which is reported. And funny enough, the founder, Alex, he had cofounded OpenSea before. And OpenSea, we all remember from that NFT era, which, of course, got all the hype by being the first and the largest marketplace for trading and holding, NFTs. And now people on Twitter are actually joking that his next something will be Open Robotics. He's basically writing all the ways possible at the very right time. But open order, I mean, my question to you, technical question, does it help you reduce cost? Like, why will anyone want, to use an API to access 400 different models from 70 different providers?

Dhruv Sharma: The right model for the right job, I think I've been using this phrase, which is, you know, the the the standard phrases don't bring a, you know, knife to a gun fight. You also don't bring a gun to a knife fight. You don't need the Frontier model, to do every single thing. The other thing is also certain models are just good at certain things. So if you need a voice model, you're likely getting it from ElevenLabs or serve them in India, and some models are better for coding. It depends on the end user application as well. You know, some are just created reading documents and so on and so forth. So it's a little mix and match, mixture of experts as that framework, is called. And, so it it does so that's that's the number one argument for bringing your your blended cost comes down as opposed to just using. And I don't know if if that's the case for you to offer our listeners. But for me, all the frontier models that I use in my AI tools have been glitching a lot. For one thing, their response takes forever. After twenty minutes of them being on a task, I'm like, hey.

Utsav Somani: Know probably the thinking more. It's everyone's using, like, these high end models to, I think, do, like, very basic stuff or could be, like, I mean, something mismatch or maybe it's just overloaded now, I think. But I've seen that I get more usage done, on ChatGPT and Codex right now than actually, Claude. I mean, the Claude plan turns out very, very fast. Alright. I mean, we can quickly talk about the SpaceX, completing its 60,000,000,000 acquisition as well, with, Cursor, which is the largest acquisition of any venture backed startup on record, and nobody got paid in cash. Cursor shares converted to $389,000,000 SpaceX, shares, and the founders and investors did not take an exit. So they basically took stock in a company that just went public. And, of course, there's so much hype around this because SpaceX is now called SpaceX AI. So they're really going after this with their Colossus cluster and a whole bunch of different things, and now they have a code editor, to go with it. And Grok also released GrokBot, which, lets you do a whole bunch of different things. Do check it out. I think you'll need the Pro Max plan, to play around with it. Alright. That's it from us. I think, we'll see you on Wednesday. Thank you so much for tuning in. Bye bye. Thank you.