The Wisdom Wall

“there's many times you would be hearing this, um, argument in our country that the farmers are looted, uh, they get Um, for, for the papaya or for the apple they produce, they get only 10% or 20% or 30% of the final price which is paid, right? This is a very short-sighted argument. Uh, and this flows from what I…”

“Our hypothesis was that to find product market fit with something very new, Most of it was luck. We felt talent and hard work came nowhere near guaranteeing you would. I mean, maybe, maybe there are some incredibly rare talents. Some people can see it, but I'm talking like 99.99% of people. We felt, um, funny power…”

“If, if, if you, if your, uh, you know, great economists are, uh, basically, uh, glorifying poverty, then don't complain that you get poverty, but because in each one of these things, an economy arises out of it, you know, so you will get, uh, an, you know, uh, uh, an entire povertarian system will emerge out of it.…”

“One of these billionaire foundations will go and fund these studies. Then those studies will show either nutrition, vitamin D deficiency, whatever, whatever it is, the latest thing that you want to sell to India. And then it will be, then next stage will be, it will come out even more sensationalized. In the press. So…”

“I think the action needs to be justifiable, uh, justifiably malicious or provably malicious.”

“Nobody can profitably deliver every order in 10 minutes. It's not possible. The network of dark stores you'll have to set up to do that is not profitable. So you will never be able to build a good unit economics based on only 10 minutes, which means all orders in 10 minutes.”

“that's the third immutable law of brand building that you can only build a premium brand and make it go mass. You can never build a mass brand and make it go premium. There is not even one example in the world where someone built a brand, which was first adopted by rural India, and then started selling in Bandra and…”

“capital doesn't follow fundamentals. Fundamentals follow capital.”

“India doesn't have a middle class boom. It, however, has a huge lower middle class boom. Right. A huge number of people entering at 20, 25,000 rupees monthly income.”

“And it's because of new behavior creation that incredibly long, large long-term value gets created in the world. And India is almost the antithesis of it. All the large-term value you're trying to create is in Me Too's. Right? By pumping more money and trying to win market share and win, you know, pole position on…”

“I believe, most people love founding stories. Most people want to help. And most people will find a way if you speak with them and tell them this is the problem you're trying to solve, they'll find a way to help you. And that doesn't mean that they are going to be the right kind of customer for you in the long term.”

“you have to like leveling structures, like you have to foundationally reimagine how you operate as a company, uh, to be truly AI native or on the frontier, uh, versus I think The other option of it is you could say, hey, this is how I do things today. Uh, I'm gonna go slap on AI a little bit on top of it. Um, and I…”

“You either never find product market fit or you find product market fit, but it's an average product market fit. And so you peter out at five million or seven million because you can't scale beyond that.”

“Well, I actually think they do. I think if you're trying to sell to the frontier, um, what, how are you going to have a conversation with a customer who's talking to you about, you know, the difference between model gateways and three harnesses and memory systems, like, how are you going to have that conversation? Why…”

“If you know you'll have three customers, I call it as a kiss of death. Because anybody like who's 42, 45, who is in an industry experience of 20 years, you'll be able to get three customers, you know, I know.”

“The clinicians had the tools they needed for the average patient to get the job done. The real problem was administrative. It was in how they ran their telehealth operations and how did, and how they did their billing and how they, you know, fought insurance companies.”

“Strong Indian universities do so much better, ah, because of like local, ah, demand and acceptance.”

“And a lot of times what founders end up doing is they say, okay, it's so easy. I just hired two sales rep in the US. And I'm like, never, it's never, ever going to work because it's not just about your sales. Your entire company is meant and built For Indian needs. The pricing, the packaging, the positioning, your org…”

“They tend to degrade in terms of their intelligence when you try to build models that can do multimodal reasoning. Like text only models tend to be pretty smart, but when you do multimodal, they tend to get a little bit less smart.”

“Now, when you're doing this kind of end-to-end thing, you are less likely to get disrupted. In fact, if, uh, you know, if anything, if the, um, if the cloud models or codex or even Gemini models or any of these other models, uh, get a lot better at, uh, coding, these people only benefit. They don't, uh, because there…”

“The number one thing before you, you know, you get into defensibilities, make sure, um, that whatever it is you're building is, uh, uh, is, uh, was impossible to build four years ago. Make sure, you know, it is a 10 X or a hundred X better than the status quo. Make sure that, uh, it has a capability to, Uh, you know,…”

“Yes, while this has done a lot to the execution part, it does not really seem to have speeded up the innovation and the invention part, because theoretically, imagine you speed up everything, then who cares? There are 10 times the number of venture-funded startups.”

“there is also, I think, limits to how much value can be added there because ultimately there is just not too much new information. If you're telling the model itself to kind of generate, yeah, There is a, I mean, there's all sorts of things about, uh, uh, how, uh, um, beyond the point it is really not going to help.”

“having, uh, having the need for a meaningful services component, uh, does, for example, uh, I mean, it serves a useful purpose also today, right? Like in a manner which might not have been, um, valued as much, uh, a few years ago, which is that, uh, It absolutely gives you a certain moat against the likes of a next…”

“I, even a couple of years ago, I always thought, uh, certain categories of startups, which were doing some, uh, almost, uh, like a bandaid on what is today's, uh, a hole in today's model is really not going to survive because the next version doesn't need that bandaid basically. Uh, like a lot of, uh, uh, yeah, like,…”

“Most of the mode is not going to be technical unless you're a foundation model company or a world model. It's not technical.”

“The foundation models aren't built for that because they don't really understand persona. They don't have memory. You know, you think they don't have memory of multiple complex and the interdependencies between those. So I actually am seeing founders, some of the better founders I see as actually combining agentic…”

“the pace at which you can get a buyer and potentially even churn is much higher, right? Because the AI products, one of the benefits of that is it is easier to build, but also I think it is easier to put it in and replace it, right?”

“companies really need to think about how are you building that full stack between the model layer and the UI. Like, you need to own that entire thing, because if you own the entire thing, at least in some verticals, It is a lot of work to rip you out. And second, if you build that well, as the models get better, your…”

“In our business, I think three things matter in that order. I think number one, it's a parallel business. So you've got to be in the best companies. A few companies end up driving the returns, whether it's for your fund or for the, for the industry, and you've got to be in those companies. Number two is ownership.…”

“And one of our core belief and thesis is that when building consumer hardware, you do not want to, with your first product, create a new market.”

“We feel like if you really focus on the fundamental problem that, hey, evaluations for language models is a very task-specific, constrained problem, and it does not require you to use, uh, you know, general purpose reasoning for that, and then there's fine-tuning you can do on the proprietary data. You can achieve…”

“in an era where you can build an, you know, a website and a web application with a Postgres database behind in 30 seconds, Software is just commoditized. Software is just tokens. Code is tokens.”

“so somebody who does their job well in terms of underwriting these institutes or the courses, the end uses and the students will have a very low NPA. It will behave like a secured product, uh, is what our understanding.”

“the effort required for doing something decently big to doing something magnanimously big is not that different.”

“So the retention in MarTech is always subpar. That is a well-known fact, which creates challenges in fundraisers and everything else.”

“the likes of Adobe, Microsoft, those companies, right? Those two, um, They don't move as much until it's like a billion in enterprise value, right? Because it's just too small for them. Um, so players like us are not attractive to them in general.”

“sustainability has to come through some sort of business mode and not through just technology mode alone. Because there's nothing called a technology mode and that's gone for everybody, right? It's become a zero sum game for all of us, and when software becomes close to zero, then the modes are going to be the business…”

“if you can solve, um, coding, tool use, reasoning, and multimodality, you have the keys to superintelligence. Uh, those are the building blocks. If you can solve those four, you can automate anything that a human does in front of a computer, and it'll give humans superpowers.”

“Basically, um, the agents create V-one of the work. Humans are verifying and iterating. Humans never create V-one.”

“So on one hand, the models are becoming agentic. And in theory, if the models have access to the right sources of data, You kind of don't need software in the middle. The models can do it end to end.”

“step one should be to, like, just do it with a frontier model and see how well it does. And most times you'll see that you don't have to touch the weights. But there are certain workflows like customer support, um, like an invoice to pay system, some of these more basic stuff where I think, you know, a smaller open…”

“there used to be like, Battery and Neeraj had mentioned whatever the three triples, two doubles kind of a thing, uh, triple, triple, double, double, double, or whatever that, whatever the norm used to be, uh, that playbook's out of the window, right? Like right now with these AI companies, as I said, like you see…”

“Now, in the AI wave, I feel if you're heads down and you're focused, you don't know what's going to come hit you from the left or right. You are one anthropic press release away from not having a business ahead of you.”

“Because the traditional model of, like, enterprise reps doesn't get you to a hundred million ARR in two years. You have to do something drastically different for that, right? In fact, a lot of these hot AI startups aren't doing the traditional model. Be it forward deploy engineers, be it selling services contracts, be…”

“So yes, there's, so instead of like, 80% of the playbook being replicable, now in the AI world, only 20% of the playbook is replicable. So I don't know if being a repeat founder has that much of an advantage in this narrow time frame where it's AI.”

“what happens in large enterprises is self-serve is a little bit of a myth. Because no senior guy in a large enterprise No one wants to sit in a two hour training and learn a UI. So you, unfortunately, you get delegated out to an agency or to a, you know, a system integrator or someone using your software on their…”

“I now strongly believe that strategy is something you do once and two, three years. Because you have to let it play.”

“in the old world, everything I just said is on people. It's on the salespeople and the delivery of customer success people. I do not accept that in the new world. Everything should be a product problem. You show me a problem in a company, I will tell you how it is a product problem. Your customer sat is wrong, product…”

“we hire, like, a lot of mid-market A's, not, we did not hire, like, your traditional enterprise A's early on... the reason for that was that we wanted to build a fast velocity motion, uh, which was, like, the, the deal size was about 50,000 dollar, and, and we wanted to do deals, like, With the, with the sales cycle…”

“what we realized is that it's the developer at the larger companies wants to use open source, and if you are a developer at a, let's say, born on the web company, you actually want more of a, um, off the shelf self-sell product. It doesn't have to be open source, and the reason for that is, uh, if I, if I am, let's…”

“And by definition, you maximize carry or you maximize multiple by making your largest investment first. And so that is very much the strategy of engineering capital. The first check is always the largest.”

“The second thing that has happened with AI is that for the first time, we have the entire technical stack being innovated on at the same time. This has never happened before. So, you know, in the PC era, all the innovation was the PC. In the mini era, all the innovation was the mini. In the internet era, it was all…”

“one of the very common failure modes of founders, especially technical founders, is what I call playing house. What I mean by playing house is that in any company, and this is related to that prioritization point, in any company, there's a large amount of work which needs to be done, which is good work. For example,…”

“in a good business, in a well-run business, you should definitely get carry checks in less than 10 years.”

“And I've heard this in conversations where a VC would be like, do you, do I invest in startups, which is like a risky business and I'm still getting a 10 X return or do I just put my money in Anthropic and I'll just get a 10 X return without doing any work, without doing any work. So totally true. And I think founders…”

“Not at all. I, I'll tell you in this, I feel like this is more coming out of experience than, because obviously as all of us AI founders, right, we, we all have this theory going in and, you know, but I would say the beauty with AI and I would say beauty with us was like, I mean, we, the results prove it, that no, it's…”

“He says, one-on-one means politics, because I say something in a different context to one person, they go off and use that context to get things done. And sometimes reality of what our conversation was and what happens is different. So he says, when I get a request for one-on-one, I said, what problem are we trying to…”

“if you have never taken something from zero to one and you've only ever actually worked in large companies, Then you are at a disadvantage in trying to navigate the early years of your own company. It's just way harder to know what good looks like.”

“You learn way more from success than you do from failure because success teaches you the things that do work as opposed to eliminating one of the 10,000 things on the list that don't work.”

“every time you find yourself doing meta work, you should have a micro panic attack because you are burning precious time on something that yields no benefit to your business. It's a necessary evil, but it should be minimized. And so you asked about like, you know, when you, when is it too much process? To me, the…”

“The PM design and engineering jobs in their most coarse grained forms, not in their like, really find detailed forms, but in their core screen forms as roles are all converging.”

“Most companies, most companies, again, set aside the outliers. Most companies that find wild success find it pretty soon, find it pretty early. They find it because they were in a good market. They find it because the founder had an earned secret that they could use to build a company that no one else could build.”

“especially as an early stage, uh, company, don't do business. Don't take money from, Funds that are in any way loss averse, because that's where they get into this sort of pernicious misalignment of incentives, um, you know, encouraging you to keep going because they don't, they don't want to mark you down in their…”

“Every startup in the AI application space, every, you know, service as software startup has to first and foremost internalize that their vendor is also their competitor. The model provider is both vendor and competitor. And that's a hard place to be. And they are, they are actually educating their vendor on how to…”

“that once you have grown to hundred million dollars in revenue by buying three companies, growing to four hundred million dollars by buying 10 more companies, that's not a network based act. That you will probably have to do some special stuff to do those things in parallel.”

“Every six months you're rediscovering product market fit, whether you're a startup or you're a large company.”

“especially when you are B to B enterprise or like selling into other companies, good products or okay products with great GTM can still win out. Great products with bad GTM can still lose out.”

“And I also warn founders saying that, hey, just because you sync first party data, you know, many people will tell you like, hey, I have the network effect because all my customers give me data and I have the network effect. That's not more because every enterprise customer will write in the contract that you can't use…”

“when founders are not financially secure, and let's say they put in five, six years in the company. They became, they become so fearful as this is a hundred percent of what they own. Yeah. This is a hundred percent of what they own. They cannot let it go because they've put in so much time. They cannot let it go to…”

“If you're building a venture-backed company, so that means that you're, you're basically kind of looking at, okay, how can I build a hundred million dollar or multi-hundred million dollar year at a company, right? Um, which means that, like, you might, you would, you would kind of like, your growth would come down, uh,…”

“For the middle class and the upper middle class who have made servants and so on, their rupee is more valuable than the dollar in terms of purchasing power parity as compared to nominal, right? But for the poor people, it's a double geo parity because they are getting little wage and they're not consuming those things.…”

“In a world where agent tech is taking decisions, or even AI is trying to, if someone's asking AI, boss, uh, please tell me some, uh, let's say pen testing tools which fit in this budget, then you have no indicator of pricing, neither to your customers, nor agent tech, which is trying to actually make decisions on, on…”

“I still believe that products are no more the moat with AI first world, but in, in, but, but really hard, in really hard startups, which were, which go beyond just products, I feel, uh, still, I mean, the problem itself, uh, becomes a moat if you are trying to solve something, uh, something hard.”

“this country still runs because 10% of politicians are fantastic people. This country still runs because 20% of bureaucrats have not sold themselves.”

“if you step away from the technology, ultimately what is going on is we're seeing the shift from co-pilots which assist a human To an autonomous agent that for a certain class of tasks can replace a human.”

“Every startup in the AI application space, every, you know, service as software startup has to first and foremost internalize that their vendor is also their competitor. The model provider is both vendor and competitor. And that's a hard place to be. And they are, they are actually educating their vendor on how to…”

“If you cannot have a product that can compete with GPT-Five, you're dead on arrival. Because you want to spend the next five years losing money. To the point that you can have a profitable business five years from now, and you'll be competing with GPT-Sev.”

“So, because of the antitrust issues, companies are choosing to, you know, lift out teams, including the IP. But the bet they're making is the IP is changing so rapidly, you take a licensed copy of the IP, you leave the, you leave a copy behind, And that team, when you acquire the team along with the license copy, in 18…”

“The error is gone, right? So now you can't do like, okay, I will actually bet on digital sales. I can able to sell as nine dollar per agent subscription. Those days are gone, right? So those should be rethought about how you wanted to build digital sales. The funnel itself is different. I'm not saying the funnel is…”

“You can actually build an application just a layer on top of a legacy systems where it can able to give you immense value, but you will be replaced because a new tool comes. So it's like there is no stickiness to it, right? So you are actually as good as a workflow where a new workflow comes, you can be better…”

“So actually voice of customer or customer feedback is no longer necessary. You can now run generative AI on all these calls and you'll automatically now know earlier you used to ask for CSAT and NPS. All of those things are not required. You can actually pretty much build your product roadmap by just listening into…”

“So you start with a hundred seats, you go to 20 seats, but you do AI conversations, your bill will now increase from, let's say you started with one lakh MRR, you'll actually go to about five lakhs MRR. So, um, so we find ourselves in a really interesting situation. It's a very exciting situation because contacts and…”

“because of our enterprise focus, we actually now have left a lot of open space in SMB. And, uh, so if you are now thinking about getting into our Actually, the way to attack Exotel is to start with SMB and then grow up and then the cycle just keeps on continuing.”

“Earlier we used to talk about two fronts. Now it's just one coordinated front with Pakistan as the lead player.”

“the secret of early stage investing is actually to invest behind founders and teams and companies that are going to succeed anyway. Maybe we made two percent difference. But you've got to be humble enough, enough to recognize that.”

“In financial systems we create, no, between subsidiaries, between group companies, we create all kinds of walls. We don't allow, uh, industries to own banks, for example. So we have to allow these kinds of walls. We have to create between different kinds of AI so that all of them don't break down at the same time.”

“a startup vendor may release features every week or every day or every hour, but enterprises cannot consume it. At the same scale and speed, because you need to change management, you need to do quality testing.”

“as per me, sweet spot is if you're able to find, let's say one or two direct competitors with some customers, That validates the problem exists, but also validates that it's just very nascent where the, if the whole market grows, your company will grow with it also. So I would ask, don't build it if you have zero…”

“It's a politically incorrect thing to say, but for a lot of product builders, their focus is as much about getting promoted at the end of the appraisal cycle as it is about building a product which really resonates with the audience. And when that is your incentive, then it is in your best interest to push out feature…”

“The synthetic human, the best thing which synthetic humans can do, which the AI agents cannot do is they can fail. They can fail at doing a task, which is a reflection of how a real human would fail in using a product. An AI agent will always figure out, or the goal is to always figure out how to solve the problem. Uh,…”

“My belief, my hypothesis is that product building or generating products is going to get more and more democratized thanks to all the advance in, uh, you know, generative AI recently. But what is not getting democratized is user understanding.”

“We have an opportunity to create Indian companies and the Indian companies should focus on the Indian market first, right? Because this is where you learn, you learn how to build products or you learn how to improve on a product, right? You know, quality performance, how to sell, how to get the supply chain done,…”

“semiconductor doesn't create, you know, many, many jobs. Because it's very highly automated, especially manufacturing is very highly automated and so on.”

“without manufacturing, right, you cannot create innovative products, you know, you know, so you need that feedback and, you know, uh, uh, you know, how, how they work together.”

“I personally have given a lot of talks to the financials committee And they're, you know, starting to open up their purse. Um, they're a little bit, uh, you know, not used to this, right? Because most of the funding goes towards salaries or, uh, to tools, right? In our business, right? So they're not used to the, and…”

“See equipment making is basically, you know, the big companies like apply materials and lamb research and all that, they're big fab equipment. They basically integrate, right? They have thousands of suppliers, thousands, and each supplier will be a small guy. And I think that is what is required in India, right? Not…”

“I think more than revenue, what I would more index on is product shape and product usage, because your revenue is your lagging metric. Sometimes you can have a lot of revenue, right? But maybe you might, you know, just cap out at like five or six mil right over there.”

“trucking in India is a very fragmented market structure, right? And you're competing with this, uh, very, uh, small scale individual trucker, right? Uh, in a environment which is not very regulated, Right. Um, that cause, that can cause, uh, uh, you know, um, issues in your, in your economics over time, because there…”

“I think it doesn't matter, uh, my personal, that's my personal opinion, it doesn't matter how, how fast you get to first million, because you're figuring out the product market, right? You might have to do one pivot, two pivot, you need to find out the right segment, right pricing point, value delivery, product…”

“which I think is a lesson that a bunch of founders and investors learned together here over the last 15 years, that you can't just aggregate at the high level and just provide access. Yeah, ok, uh, air tickets and railway tickets, fine, because those are pretty commodity sort of items, right? But a stay in a specific…”

“You start off, you don't go into a huge negative cash flow. You're going to shallow Jacob and you come out, but you don't grow at, you know, there's no hockey stick. Three X a year. You grow at 70% a year, 50% a year, but you do that for 15 years. If you go 50% a year for even 10 years, you're going to be a large…”

“And these markets are much bigger than software markets.”

“my view is that M&A really gets going when you have multiple companies that are worth more than twenty billion dollars that are, that end up being buyers. And we don't have that. Maybe Zumaito is the first one. There might be a couple of others in tech that get there, but that's it.”

“it'll take you one or two years of wandering around in the US to figure out the exact wedge. Um, you know that one of your founders will have to move there. You know that you will churn through your first VP sales. You, there are these things that are kind of like, it will happen.”

“Because investors know India and they have generally mixed to negative views because venture capital has had liquidity issues. Private equity has had high valuation issues. Actually the people who have consistently made money have made money in the public markets, right?”

“today, you know, it's safe to say that if you have Economic power and military power, but no technology power. It doesn't count for a lot. And if you have technology power, you'll be able to accumulate the other two parts relatively easily.”

“They, they think every open source project is a DPI. They think every government tech is DPI, which is not true. Uh, some government tech is DPI. If it enables private entrepreneurship or private innovation to happen on type on top, some open sources DPI. And all DPI is not open source.”

“by looking at the bookends on the distribution side and at the data side, We may allow for a fairly liberal system, uh, in the middle.”

“today the pace at which technology is going, it's very hard to have 12, 24, 36 month lead up in the market. Everybody has an idea, right? That whole execution of that idea will make the biggest difference. Uh, and if execution becomes your biggest moat, then your distribution channel becomes the most, uh, important…”

“LPs as well as GPS. We have all realized that all numbers are good, but there's only one number that eventually matters. And that is DPI, right?”

“I think they're okay taking sometimes even lack of performance in a particular fund, but they're backing a certain strategy. What they don't like is sort of macho behavior where you said I will do X, but you end up doing Y because then that's not consistent with the, with the exposure that they are taking for, uh, for…”

“Then what will Flipkart and Amazon do? Wait till the category becomes large enough. Then they will open one more tab and you will be wiped out. And we have seen enough examples of that happening across fashion, across, uh, baby, uh, products.”

“The effort involved in collecting one rupee, I would argue the same as the effort involved in collecting hundred rupees.”

“I think people are a little bit, um, Naive about the downsides of raising money. Um, ultimately your investors will want an exit and it may not be at the same time as you. Um, and for a business that's as mission driven as chess.com is, it's, it has at times been irritating to me to, to have investors that feel like…”

“So, um, you know, kind of that, that certainly can always be fixed, uh, over time and, and kind of managers can learn proper portfolio construction. What can't be fixed is your taste for, um, entrepreneurs and for unique opportunities that can become these outlier opportunities.”

“Lots of second time openers that were able to raise lots of money very quickly, um, came down with a top down thesis on, on a market. Right. And, um, Never really got to product market fit because the team that they were able to build, um, because the capital they could, could raise was way too large to get to product…”

“India is the only large economy left in the world, which has, as I see it, four parts to its economy. There is a Bahrain on top of a UK, on top of a Brazil, on top of the entire African continent.”

“keeping an undervalued exchange rate is very often good for the economy. This is a very important East Asian lesson.”

“consultants are trained to tell the management what it wants to hear. Very rarely does a consultant make a recommendation to a top management Which actually would make the top management realize that, you know, what its prejudices were wrong.”

“Kolkata should have been that growth zone. Anchor of the East. So bringing back Kolkata is not just about, ah, reviving West Bengal. It is an anchor of the East, because when it gets going, it, it's just overwhelmingly bigger than Bhuvaneshwar or Ranchi or Gohati or Patna. Right. It'll take us a very long time to…”

“there will be a time when, uh, certain Indian states will be too expensive to do certain kinds of activities. We should not get emotional about it. Hand those over to another state. There are poorer states in, in queue. Give it to them. You move up the value chain.”

“If there is a person who's corrupt, then his or her resources earned through corrupt means is not kept in cash, right? It'll be him deployed somewhere. Because you don't keep cash, I mean, cash depreciates in value. So the idea that, you know, all the corruption is accumulated in cash somewhere is a silly idea, is a…”

“So this, Top 30 million market of India, which is thirty million households, which most people call India one, or what do you want to call it? It is very, very international. And this is not a value based market. Grocery, if you look at it, the popular definition is gross is value waves. People look at prices and buy,…”

“So the funny thing that happened in India was as stores became older, they became more unprofitable instead of becoming profitable, which is the reverse, which happens everywhere else in the world. Because you amortize your costs, stores become more profitable. In India, they become more unprofitable, particularly in…”

“We chose to build a Benny spoons in the way I mentioned, uh, so outsourcing, finding product market fit to the market. There's plenty of people looking to do that. Plenty of VCs and angel investors investing so that bets are made, uh, and rather Benny Spoon's coming in afterward to take something good and try to make…”

“AI makes capable people more capable. I, I, unfortunately, perhaps It increases all sorts of divides, bearing governments spreading, you know, opportunity and wealth. But, uh, in the wild, it increases divides. It makes a great engineer better by 10 times. And the mediocre one who was already one-tenth the ability, if…”

“If an average sales guy is able to sell the product without in a lesser time than the previous best sales guy, I have achieved PMS because this is pulled from the customer and average sales guy should be able to set a very, uh, non quantitative way.”

“See, the most hated products in a company are the most difficult to replace.”

“while sitting in San Francisco, uh, you know, you might have felt like, oh my God, this is a sweatshop for that person. It was upward social mobility. And as much as they may not be happy with that existence, it meant that their life was improving.”

“often, you know, when we discuss India and China, people say India has too much democracy and too much politics, and China functions better because China does not have politics, right? It's a single party state, central control. That's not true. It is a single party state with strong central control, but, uh,…”

“Likewise, the companies that came in with joint ventures sooner or later, their joint venture partners Got the skills, understood the technology, understood the processes, and broke away, and then were preferred by local governments for their subsidies and their support systems. So then they became competitors of their…”

“Instead you use state capital on what you see as strategically important domains, semiconductors, strategically important. You need certain kinds of dependencies to go. There may be others. There may be critical minerals where you support. You don't need to do it for everything. Why do you need a PLI scheme for…”

“The bit where Silicon Valley starts to have a pretty overwhelming advantage is if you're going from a thousand to 10,000. There are just not that many people outside Silicon Valley who have run thousand person engineering teams. Um, but there are actually lots of people in Silicon Valley who have done that. There's not…”

“So your co-founder has to be meritocratic in nature instead of your, the first friend from school Uh, or your first love from your college coming and beginning to work with you. I think that's what I'm calling manufacturing co-friends, which means co-founders, which, which means, you know, that if I'm starting in AI,…”

“for, uh, seed and pre-seed, we've taken the approach of, we think, uh, smaller managers are better suited, um, to invest in seed and pre-seed. Uh, we, you know, we have a long tail of 25, you know, ish, uh, firms in, in that, uh, and most of those we've added in the last sort of three years. Um, and the, when I, when I…”

“One is that in any game of skill, there is a learning curve. That is, the more time you spend on the game, your skill should improve. Uh, so that is, again, uh, all of these has to be static correlated. Second is persistence of skill. That is, if you've played enough number of games across two seasons to Performance…”

“because of the power law dynamic, unless you're A winner, a clear winner. They're actually not incentivized to work with you, which is the irony is like, okay, the person that needs the most help, they're least incentivized to help.”

“The way we're seeing AI being incorporated into a lot of businesses is that it's more of a sustaining innovation and not a disruptive innovation, if you were to use kind of Clay Christensen's framework. In, for example, in BtoB, it hasn't become a big enough wedge yet to build an AI native app company that solves a…”

“So in my view, healthcare is a patient capital. Healthcare is long duration. Healthcare is not a rocket ship. Healthcare is, uh, grows, becomes steady, and then continue to provide dividends in that kind of an industry. And the kind of capital that needs to go after it is more the PE rather than necessarily the VC.”

“that is entirely driven by, by the assets, uh, particularly because at that point in time, the GP Can't screw it up either way. Uh, you know, there's not really much left to decide on. Reserves have been mostly deployed. Um, it, the, the GP itself doesn't matter as much.”

“In emerging markets, consumers are not willing to trust companies, and so you have to build a brand to a certain, um, size and scale, and then they'll start, um, enabling, uh, trust to be able to try multiple products. It's why you see so many companies like a la Zomato, not in just one business, but three businesses,…”

“If you are going to take them and get them to run PSUs, they'll obviously not be able to spend energy time thinking on healthcare, education, judiciary, where we have a big problem, law and order in general. So all the developmental arms of the Indian state are anemic, right? They don't have the capacity.”

“So at the lowest income level, right, they have disproportionate influence on who gets elected. Okay. At the middle in, uh, income, uh, strata, they have disproportionate influence on what gets discussed, right? Because you and I, the people who are doing podcasts, reading op-eds and things, and at the upper, uh, level…”

“So the initial fear was if you introduce capital into this, the employment will reduce, but the learning was actually that, you know, young firms create employment. Uh, not small firms. So the, once you de-reserve many young firms came, many firms came, and then the overall employment in that sector actually increased…”

“what happened in the 19 nineties was that because the government's Where coalition governments, they were not strong governments, so they could probably experiment with many economic reforms, but from the 2000 onwards, uh, governments, uh, have been, 10 year duration. You know, the UPA was 10 years, then NDA has…”

“You need about not just Adani and Ambani. You need 10,000 large companies in India because they will create the ecosystem for MSMEs. They will create SMEs. There'll be backward forward linkages, and that is large companies are then able to penetrate global markets.”

“So, uh, you know, with our four trillion market cap, our capacity to absorb companies every year is probably about, you know, say two to two and a half percent of the market cap, which is anywhere between say, 80 to a hundred odd billion dollars.”

“So I feel the people, the problem with AI is it makes the people with the resources even more powerful, right? Because it automates the, the process that you used to do on an everyday basis. So if you have some leverage that is not directly related to your time”

“I think the fundamental innovation with EdTech is not technology. It is CAC. It is destruction of CAC.”

“I think one of the, one of the problems of, uh, say a place like Bihar is not that it had bad leaders, that bad leaders are a reflection of what that society aspires for.”

“A 30 to 40% gross margin product cannot sustain through a website. You need a 70% gross margin product. So just because everybody does D to C, Personal care can do D to C. Food cannot, because your gross margins has to be north of 60%, because you have the logistics cost, you have the marketing cost, then you have to…”

“As a shareholder of a bank, I will say that banks should always take equity in case they may forgive. Even one rupee of a loan. If I forgive one rupee, I should have equity against it.”

“That free electricity came from a political decision of giving to farmers subsidized fertilizer and subsidized electricity for groundwater. That decision could not be taken back by any political party. So again, I would believe it's the negative dividend of democracy and our five-year competitive democratic politics…”

“Some people say that South Korea and even China moved fast because they focused on exports. That is true. But that is no longer possible when even the developed countries are not going to grow fast. And the attitude in the developed countries is also changing in terms of free trade. They taught us about the advantages…”

“You are in a job which is the highest level of intellectual sales.”

“They had no working vocabulary of English, they did not know the language, yet when given an option, they will pick that language. Because of aspirational reasons. And in fact, there was a lot of trust deficit. A lot of these regional language apps in the past had always been these betting scams and all of these…”

“this narrative that tier one is where all the disposable income is and tier two, tier three is where you don't have. Is not the truth. If you have a very widely distributed product in every single city, you will find that top five percent with, with all the disposable income.”

“You sleep for five hours. Your recovery in the morning is going to be almost equal to sleeping in six hours if you can trick your mind. And the trick in the mind is don't tell him how many hours you slept.”

“And the other thing we're seeing is at least one founder who is involved in go to market at some point in time, early in the company cycle, Has to spend time either in New York or in San Francisco. There is no alternative for that.”

“But if you're going to make me grow this 80, a hundred, a hundred percent year on year, I don't think it's possible. Um, I think it just puts very wrong measures and we all face it, right? Every time we have to go raise capital and it is important. You have these weird like metrics that you have to chase. I think it…”

“Dairy farmer in India never makes money. It's a cash flow system for him. He thinks he's making money because it doesn't count his labor. It doesn't count his son's labor. It doesn't count his wife's labor. That's the problem.”

“In India, you get actually the ability to influence because you have taken this approach of being no greed, no glory. So you can now say I'm genuinely doing it for this cause. There is no hidden agenda that I have, which is either to get glory or to make money. And so you are able to co-opt others in the process.”

“Today, tech is all about S curve inflection. After the first S curve inflection happens, it's already too late to enter the market. Not only for the entrepreneur, but even for the early stage investor. I say to early stage investors that it's already come in times of India, it's already too late for you, right?”

“For a investor or an entrepreneur, This is a good audience because this then again, like I said earlier, The expenditure people will do is not restricted by the, the individual's income. It is about the income of the individual and the parents.”

“particularly when the business needs a three year, four year discovery phase for PMF, then being too sought after by investors actually works against you.”

“Most founders, when they dilute too much, they become irrelevant and get kicked out of their companies in 10 years, which happened to some of the names that we mentioned, and then they don't get the remaining 40 X because you know, the real juice is when you become profitable, when you list, you already are established…”

“And how, how much capital in a lifetime, maybe in India, you probably don't want to raise more than fifty million capital ever.”

“If you have 50% investment from a company, then that company becomes your exit. Unless that company is an investment company.”

“So in a private market transaction, you may not look at the things in the same angle. You don't look at conflict of interest in the same way public markets do”

“public market valuations will always be lower just to compensate for that effect unless your company is so good that they are saying ki, you know, I love it.”

“My funda is, in fact, a VC should go and tell companies, please list on the SMA exchange. I will buy your shares on the SMA exchange. After that, if there is no liquidity, it will come one day. Your e-shops, your, your employees can sell on the, this thing exchange.”

“generally what acquirers will do is they, they try to structure the deal in a way that The initial payout is not very high so that, you know, the founders can actually take most of the money after six months or a year, right? Instead of you giving that money to the investors, right? Because the investors only get the…”

“because you, you feel like, okay, I can't do, like, I have to do a really great job with everything, you know, but in startups, you know, the velocity matters. Because if you, if you have velocity, you can iterate faster, right? Uh, this is also like one of the, you know, the curse of experience, right? And the curse…”

“for second time entrepreneurs, um, it is difficult in a different way, wherein I think they, because more often than not, they are financially free. So they paint a very idealistic picture of what they want to build. Uh, almost like MBA, this is a large market opportunity, will, they would have a long plan, step one,…”

“We create one and a half crore jobs a year. But 80, 75, 80% pay less than 25,000 rupees a month. So job creation is not a problem in this country. The problem is the wages.”

“Now, SaaS is the phase that has to come, but for SaaS to succeed is not technology ability that's important. It is the ability to get capital, ability to market and sell, and ability to get a reasonable fee.”

“See, we Indians have not learned the art of marketing. An American kid will sell a 25 dollar stuff for a hundred dollars and make the customer feel he's got a great deal. An Indian will sell a hundred dollar stuff for 25 dollars and feel sorry for the customer.”

“What like in our own journey, we've seen that it's the biggest mistake is to try to build a business around what will be cool for a VC, because there is no way you can be the smartest person in the world, but to predict that, you know, at that time where, you know, what flavor of the season, uh, because if you start…”

“once you build hardware, once you've stabilized your platform, you've got a really powerful lead. You've got a really powerful moat that internet businesses can never deliver on. Almost never. SaaS maybe. Internet powered businesses, almost never.”

“It's best to build a muscle in the company yourselves first and fail a little bit at it before you bring in the real leadership and before you bring in the big guns. Every time we've attempted something as a lazy approach, let's just hire a solid VP for this and get out of it. It's generally not worked.”

“We are not a country where you can sell stuff on Twitter to a meaningful audience. You can sell stuff to Twitter for like five crore people, but tool is about by like 30 crore people. So you have to, you have to, you should almost consider, which actually we've not been successful yet at, but you almost need a washing…”

“I love newspapers. Front page ads. Nothing works better. Like the amount of football they bring in, I kid you not. They pay themselves back over a weekend. Like the best ROI.”

“What I'm saying is that when new markets have been created, uh, when industry is being restructured, what you thought was TAM may not be the TAM.”

“You know, if you look at, uh, Strategic sales. And you leave out Flipkart, they typically don't give you the kind of return that a VC investor would expect and hope for.”

“Now, as a loss making company go public, you better be sure that you're going to be making a profit in one or two years after you go public. No, I don't think public markets are willing to tolerate losses over five years, seven years further.”

“you are really not worried or, you know, it's not that important to exactly pinpoint with Six Sigma accuracy on what is this cash flow? Is it 25,000 or 26,330 thousand rupees? You know, it doesn't matter. You have to be in the ballpark because then you have enough margin of safety in the form of a debt burden ratio.…”

“You have to either choose a US or choose Southeast Asia. You can't do both because if your product market fit is good, the best company will still take the biggest market pretty quickly, right? So being in two different markets actually doesn't help because these two markets don't talk to each other. They have nothing…”

“often time, like it, even that may sound harsh, like often time a product market in India is a bad news. It's a false indicator of success. At least as of today, I think India or many markets are not easily replicable to the US. So the learnings you have, you have to unlearn them.”

“Public market anyways give lower valuation than private market. And if the peak of private market, if you raise at the premium valuation, you can't do IPO for five years then.”

“do not be scared of building like field sales. Like a lot of people in the beginning told me that, you know, don't do field sales. And a lot of SaaS founders, by the way, hate field sales, right? Like they don't like it. Uh, SaaS is usually thought of as a clean business to run an India running an FOS. So I, I think as…”

“never let the customer write your product roadmap. Uh, let the customer explain the needs and you find out what is the right way to solve it, right?”

“But I think it's always Smart to actually start with something smaller, whether you start with startups and go to mid-market, or they start from mid-market, go to the enterprise, and before ending up the largest enterprise, because these Larger companies, they're also, by definition, they're paid not to take risks, and…”

“Grocery, you can't do without owning inventory, right? That's the model that, that runs. People have tried different models. None of them have worked.”

“The other more important reason that we did it is that we have reached a stage in our company at a billion dollars of sales. You need to now get strategics or you need to get piece because you at one point go out of this whole BC world because of the quantum of money that you probably need, the amount of stake they…”

“Market leadership in B to C or B to B, whether it is B to B to B, I don't think it makes a difference. If you do not have the advantage of basically eliminating completely the competition and taking all the profits away, which means become a monopoly like IRCTC. If you don't have that, then that market leadership is…”

“one of the strange things about, uh, running a 40% growth company is that you can't shut it down and go home. Because it's still growing, uh, customers are still buying, you're, you still know that you're adding value. Uh, you're even making profits. So why would you just like, uh, throw this, uh, you know, good…”

“new trends always emerge during the fag end of a bear market, and that is why investors should always be alert to which stocks and sectors are actually breaking out to all-time highs during a bear market, because the market itself will tell you where the next leg of winners And for the next bull cycle will come from.”

“the relationships with farmers are not purely transactional. Right. So local aggregator, uh, is essentially, uh, the guy who would, you know, in Hindi, we say, Mayad Bhagwat, Shadi, Tino, who he finance karda hai. Right. So it's, it's not a clear, um, equation that I pay higher prices. So I'm giving more value to the…”

“So it's the inefficiency comes from trust deficit rather than from information symmetry.”

“You have to be clear on why you want to move to the US, right? It is definitely not for sales. Uh, America is a market that does not want to see you in person, right? They're happy to buy products online, uh, through Zoom, through, you know, like they,, they're absolutely happy to swipe the credit card on a product…”

“We say that the moment we reduce the price, we become profitable. So it has always worked for us. Like we grow, we become profit. It's a, it's a cycle. Like you become more and more efficient as you keep on getting more and more customers.”

“I think Amazon Flipkart are the true VCs. I would say they have created hundreds of entrepreneurs, like both, look at both, look at Wakefit, look at Mama, look at so many brands out there. Amazon has created those so many D to C brands, right?”

“And it is not that you always need to be VC funded or a VC backed company to get listed or get that kind of support. If your fundamentals and if your conviction about your products are clear and you know what you're building, which has a value.”

“And investors try to keep that, keep founders and early teams in poverty for longer. They feel like that person will put in Jeejaan for their cost. It's a bad idea because your burnout risk is high.”

“Indian enterprises do not like productivity tools, do not like things that save time. They only like things which make more money for them, right?”

“You are going to win by deploying at scale. You launch at such a large scale that there is no one who can then kind of even come in, uh, compete with you. And this is what, you know, really Distinguishes, you know, uh, the reliance way of doing things as compared to anyone else. But for them, everything yet you can win…”

“our view is, uh, we, on the debt side, if we can avoid litigation, NCLT, bankruptcy, et cetera, Uh, to do settlements as much as we can, to do restructuring if required, give people, founders more time. Uh, if they're good founders, if you haven't funded a, a fraud, uh, you will get your money, you know, 99 out of a…”

“you can't invest in a company like share chat by getting into a consensus, right? So it's impossible, right? Because there was hardly anything there, right? Uh, to say that you will basically create a, uh, social media company when you have a Facebook, when you have a Twitter in the country, which has millions of users…”

“if I show you right now, if I show you the world's best ad to shift a house, will you shift a house? Because there's a one by 12th chance that if you shift a house every year, one by 12th chance that you will be in that month today. So what is more important is that I need to be in your memory for a longer time…”

“The zero to one, the founder's role is the product. They shouldn't, you should not hire a product manager in zero to one. You are the, the founder is the owner of the product, right? One to hundred, the founder's role is business system scaling, right? Your whole system should be instrumented and architected so that…”

“The thing is that in general, in life, some of the absolute best things, like this is only applicable to power law driven outcomes, that some of the absolute best things, like the best hires, the best companies, uh, the best experiments, the best features are almost always competitive, which means that consensus is…”

“And that is why I believe that if you have minimum ownership requirement, and hence you will lose out on a winner because of that, then you are, even at hundred million pre, there is an upside potential of hundred x now, which was, which didn't exist in 10 years back when Eileen Lee sort of defined what Unicorn is.”

“So I think the new playbook, in my opinion, where the depth is going to be very crazily deep is just whenever you find a rocket ship, whatever it takes, you get into that boat or you get into that rocket ship and ride it through. And that is what will matter most.”

“The problem with that is, you know, it's, it's not really a lending business. It's actually a collections business.”

“So what ends up happening, you know, so that is people want to invest in NBFCs, but you don't realize that you'll be like 90% diluted, uh, by the time you build a billion dollar book, because you will just be pretty much operating at two to three X leverage.”

“If you want to come and invest in a lending business as a technology business, your valuation is just some random technology valuation, which is fine, right? But if you build that business consistently, the penultimate investor before an IPO will value it like an NBFC only. They may give you five times book value, six…”

“I like tell people often that if you don't monetize your hobby, it will die. I mean, unless if you are a good painter, if you're not able to sell your paintings, you're a good, uh, musician, if you're not able to create, uh, I mean, monetize that, those, uh, stuff that you create, it will die its natural death, and…”

“I definitely think that for most of the companies like Postman for us, ARR is a lagging indicator. Like we, among all the metrics that drives the company, I will definitely say that in, in most cases, it's not, it's not the ARR, which is the main goal.”

“We did not feel that this can be a marketplace platform because parents have no way of figuring out who's a good teacher. It is very hard to, for parent, you know, in unlike, let's say e-commerce where you buy a shirt, if you don't like it, maybe you don't wear it too much. Here the currency is student's time. So it…”

“That's probably the worst thing. And, you know, VCs do is try to design the company for the next round of financing. Versus long term, you know, tenure success. Yes, I mean, you're really the next round. But you know, smart people, smart investors figure out what is the, what is the kind of company they want to invest…”

“I like to say data is also the new asbestos. In other words, data is toxic, and data will kill your company if you're not careful how you manage the data in your organization.”

“Initially, till that first, maybe even a million dollars, right? Like, it should all be done by the founders. Um, and if the founders themselves don't have the skill set to do it, then either they should learn it, or they should bring another founder, co-founder on board, who will be as incentivized as them, ah, to…”

“Most of the people, they think, you know, my product is my mode, my data is my mode. Actually, if, and this is something that A-sixteen Z wrote about also recently, that your mode is your GTM. Your mode is like the channels that you build, the sort of capabilities that you build on going to market, rather than any data…”

“the best distribution strategy is no distribution strategy, which is, you know, let your users tell you what they love, and then build that product for that love.”

“Should not ever take an investors know as validation. That's the worst thing that you can ever do to a business because most investors look retrospectively. So they are looking at businesses that have been already proven. If you take validation from investors, you'll never be able to do anything new, right?”

“So, uh, if you think about hundred companies that get venture funded, I think maybe 10 survive. And, uh, I'm speaking anecdotally here, but I think venture capital leads to a lot more, uh, startup death than it leads to startup survival, and that is what the model is optimized for.”

“To win a market, software is not going to be the only thing that will work. We keep talking about, at least in logistics and SME, we talk about full stack solutions where you have to own the journey of the buyer because it's an unsophisticated buyer. If you throw a software over the fence, they will just not know what…”

“brand will give you stronger unit economics, but brands will have very high, ah, cost of customer acquisition and repeat rates. So very quickly you will find that if you have very high cost of customer acquisitions, Then you can't be a single product company. And if you can't be a single product company, you will have…”

“Uh, like I said, right, I mean, I am simply buying a lottery because I have no ability to move that company, right? Uh, the winner is not the smartest guy or the best product. The winner is the most well-funded there.”

“Otherwise there will be flare ups and blow ups because these companies can't raise money because they don't have an economic model. So there's a lot of subsidization of, of the customer, which is happening right now by burning through a lot of money. And that's not a sustainable thing, especially in spaces like B to B…”

“if you don't raise too much money, even a mid-sized or a relatively small exit can be very meaningful to everyone.”

“A lot of companies and products which are 20, 30, 50% better. I'm not saying that's a bad thing. That's bettering the state of the art, but that's better for an incumbent. That's better for somebody who already has distribution, already has cash flow, already has a balance sheet. If you're a new person or a new company…”

“We don't want to be shooting for less than five X funds. In India, things take time, so I have to adjust for a time factor that things will take two years more, so that hits your IRR. We're measured by IRR. And so, don't want to take an easy way out and say, three X chalega. No, it won't work. I think we have to shoot…”

“So the more common themes that get picked are dangerous in some sense for a small player like us, because then you're always under headwinds, because there's a bigger guy who says, you know, I've already made a bet in the space or adjacent space. I won't fund anything of this nature.”

“Those risky bets are the ones which are going to make the most money. Like my LPs tell me, historically, you were shying away from that. It's probably true. It was basically, you didn't know what's going to work, what's going to take off, so you're hedging, making more bets. Now they're saying, no, it's time to take…”

“Rebuilding technology is actually the easiest part of this problem, right? Like software itself is fungible. It has always been fungible. Like, I don't know why people now in AI suddenly are like, oh my God, it was always fungible. It was never that hard to But building software that people use is the hard part.”

“investing in the Orchestration layer or the model layer, like that's the easier problem to solve. Building an agent is easy. How do you actually make it accurate? How do you actually manage your, your context? That's going to be the hard problem.”

“in today's world, you can't be a leader And not be pushing your team. I truly believe like you have to be in the details and the bar is so low. There's no excuse. It's so easy. Like there's no excuse to not set up and build yourself, uh, today.”

“And then also for enterprise space, you know, if you look at Google, Facebook, Intel, all of these were startup at some day in the Valley. So the risk taking capacity of the enterprises or the customers, early adopter customers is also high. So that's where it's a combination of things why, um, Silicon Valley get more,…”

“when you find someone who actually is addicted to pain, they tend to perform really well at startups and, um, uh, and they tend to make great leaders. They tend to grow very quickly. And our thesis has always been find these people and make them celebrities in the company.”

“my recommendation to the early stage founders is, unless you have that distribution advantage, healthcare is still actually a very hard market to get into, and you're better off either working with a larger company or distributing through a larger company or cultivating that relationship somehow, because if you go into…”

“The, the issue in some ways, though, is if you're a one trick pony, you're not going to get the biggest customers in the country. The biggest customers and most important customers in the country want a platform partner.”

“And I think a lot of people think of education as a content play. Uh, but it is definitely not. Content is the easiest challenge to solve for. The biggest challenge to solve for is engagement and motivation.”

“Everything is moving so fast that if you actually sit idle, uh, you will basically be Uh, you will lose your PMF within like a quarter, right? Um, so I do feel that that velocity and that sense of urgency is one of the most important traits in AI as compared to, let's say, if I, you know, move the clock back five,…”

“the difference between a company with a half a billion outcome and a ten billion outcome is just purely the hunger, ambition, and fearlessness of the founder.”

“unless you are a deep infrastructure company where the CEO could be very, very technical, In an application stack company, the CEO needs to be a great marketer, great seller, great visionary, great product person, know about engineering, know about everything. There's, there's no two ways about it.”

“For a lot of India founders, I will say don't sell in India. U.S. is the biggest market for almost every industry. Go and start selling from here. Do not even think about starting from India. You will get stuck. You won't be able to make that jump. A lot of founders think I'll start in India and then make the jump.…”

“today when you build a voice agent, it is actually not running a 30 minute conversation, truly. It is actually running a series of 10:02 conversations, because it is basically splitting the 30 minutes into small turns, And then every turn is treated as one thing to do, and, ah, it's engineered, right? Every turn is…”

“I think the thing that, that is hard is the information asymmetry that exists, which is that model developers see where the future is before the application developers do. So that asymmetry is impossible to sort of, like, actually remove.”

“This is something that a lot of people think is mutually exclusive, right? People think either you can be a research lab or you're a product company. Yeah. We, I think what we did was we said, no, no, no, that's not, we don't think those two things are not possible to do together.”

“The one thing I think that makes this business a bit of a moving target is that the models do learn pretty rapidly. Like if you are offering a certain type of, um, uh, this thing, uh, a certain kind of a collaboration to labs with a certain, let's say, type of talent and a certain type of task that is going to get…”

“coding in general, uh, has nice spillover effects in terms of improving reasoning capabilities of models as a whole. Whereas, uh, you know, let us say, uh, the model companies prioritized, uh, something, uh, like, uh, the models getting better at medicine. I'm not too sure if that would help improve reasoning as a…”

“if the human can do at least one thing better than the model, yes, they are going to add value. Uh, that is where the costs will go. If all the other costs will go to zero, this will become, you know, uh, at least as far as, uh, um, Um, micro economics goes the, uh, this is where the, uh, yeah, this is where the cost…”

“while they were definitely data annotation companies that had, uh, relatively more scale, uh, low skill labelers. Yes, which is what a lot of the traditional machine learning pipelines needed. Um, the, um, what, uh, these, um, LLMs and the modern foundation models needed was just something very different because they…”

“The, this method is the one, I think, which has a lot of legs in the, particularly the more you can operate in this particular paradigm of prompt and then these rule or rubric based verifier tuples. The, that is a very nice way for sort of creating synthetic data, but where the human feedback was a crucial element.”

“generally speaking, there's less of a room for simpler tasks. The more, uh, um, you're doing expert level tasks, uh, the more room might be there. The, the more you are into some, uh, uh, you get closer and closer to these long horizon tasks. The kind of thing that would take a human one day or one week to complete…”

“if it's an application startup, I don't think you even need an ML researcher, um, to begin with, right? And even an ML researcher will get frustrated.”

“If you're incremental of what is already there and you have an FD model and, you know, bad cost structure, you, you literally have the worst of all worlds basically.”

“whenever a country's per capita income has crossed 2000 dollars, That economy, that's the inflection point for the economy, because what happens is still 2000 dollars per person, that society is in a survival mode. Whatever you're earning, you are just kind of surviving. There is no surplus income left to, for other,…”

“It's only in an owner-driven organic setup like ours, you have the independence, uh, to do the right thing for the business, and for the business, when I say it's for two stakeholders, right things for two stakeholders. One is the customer, and second is people who take care of the customers.”

“In trading, it's only the platform which makes money. Most investors, most traders don't make money, right? So it's like in gambling, the house is always in the money because it's always taking, you know, a fraction of using a toll or a user charge for using the platform, right?”

“radical transparency actually isn't a risk. It's actually an unburdening of your stress.”

“Don't worry that somebody is going to leave and go to your competitor and take that information. By the time information is useful to that competitor, you will have done something else.”

“I think generalists, the ability to hire generalists who can, who can harness this technology to do things is better. In the old days, you would have to hire somebody very specific specialists. Today, a generalist can learn a lot of things.”

“I think if you don't have a strong technical founder on the team, it's very difficult to build these because as you know, pushing these agents into production is, is far different than building a POC. The non-technical person can build the POC and get it to look good and validate, but when it comes time to actually…”

“by being first, multi-party, highly complex data that that hidden from the outside world, those are three pretty good criteria for, for a starting point.”

“The one thing I would say to founders is look out for your employees because chances are, especially for a lot of young founders, they're going to work with these people again. And so treat them really well. Treat them better than you treat yourself. Advocate for them. Make sure that the, that the, um, acquirer treats…”

“this is one of the few cases where radical transparency doesn't work is because it's distracting and people start to wonder what does it mean for me? And you don't actually know the answer. So there's no point in giving them radical transparency when you don't actually know. So my advice to founders is keep the circle…”

“in reality, it's really a sales job, right? You're selling people to take a meeting with you. You're selling people to take your money”

“the moment you become multi-stage, your focus shifts, and you're not able to do justice to that zero to one journey”

“The, the core of that was not key. Oh, he has a different skill set and somebody else. It's trust. And that, you know, that is how these things come together.”

“The, if you're really logical about things, you're not an entrepreneur. Yeah. If you are like one of those who's, you know, doing this SWOT analysis, or doing a two by two, and you know, and analyzing and paralyzing, you're not an entrepreneur. Because an entrepreneur is effectively looking at an opportunity, not with…”

“I actually think it's a more capital efficient model if you actually build your scale elsewhere and then use the momentum and get into the US.”

“events is not a horizontal business, right? Every type of event is different. Uh, so you guys should focus on one vertical.”

“And I said, Hey, it's just the founders who are going to define it. It's not going to be a discussion because a company is kind of built in the image of the founders.”

“Software, which if you look at software, right? Alternatives are not a problem in software. So then what will people anchor on? They'll anchor on brand, right? Because product is, is commoditized.”

“See, in the role that I play when I, even when we were building the software or whatever, gender should never come into play. If gender comes into play, it becomes very, Difficult to do.”

“If you look at it as in India, we, we value money more than time.”

“That's where I think boards and investors and as an external parties can be really good advisors, not because they know the business, because they are outside the business. The emotions are not connected to the business, where you're able to bring very unemotional thought process and get things done.”

“if I could only build one and actually build the operations and I could only build two, I'd build operations and, and professional. And then I would build customer app third. Reason being, um, like, Your customer app, um, well, it is a nice to have, um, but you can figure out workarounds if you don't have it. Um, your…”

“Uh, legacy SaaS software also applies to legacy cybersecurity products too.”

“So, overall, like, listen, investors and capital follows where Entrepreneurs or companies are doing well. So if you build a meaningful company and you're doing well, you can build that anywhere in the world. And as long as you're building it well, like investors will follow. So I don't, people, you don't need to pick a…”

“So basically the traditional like swim lanes have all been abandoned. And everyone is just all over the place and you never know who's going to enter in your category.”

“if you don't take this offer, don't assume the next offer is six months out. Yeah. Assume for the next five years, seven years, you're basically signing up for the next five, seven years of this journey saying, if I don't sell right now, assuming it's a very good and lucrative offer, then, and if you're not taking it,…”

“I would say enterprises don't optimize for accuracy of a model, right? So they're more so optimizing for reliability, um, and trust. So it's very, um, it's very, I think, in today's age with this new, ah, level of capability that the models have, getting the accuracy right is available to everyone. Ah, but who gets the…”

“Until then, I wouldn't say, don't even pretend that you're building a product because that'll hurt you more than it'll help you. Because you're basically thinking that you're building product, the cycles go there, but then they'll really just start to look like multiple products and not just one single product.”

“I think the conviction is, uh, in terms of the problem definition, in terms of the market opportunity, and in terms of, um, that I can actually do this. I think those are the three things that are needed for a deep tech startup. Um, I don't, I did not even mention market size or, ah, product market fit and all that. I…”

“I think the, um, the really interesting inflection we've seen over the last three or four months is that the models have gotten good enough that agents can actually run the entire workflow, right? They're not additive to individuals. They are, um, they're the ones that own the work and then we are additive to its work.”

“US on the other hand is, can you discover what the pain point of that individual is? US is a very consultative sales model. So, unless I have a pain point, even if you're coming to me and saying, you know, you will make three X the margins or five X the margins of what you're paying me. Mostly that customer will say…”

“in startups, especially if a, if an individual is not delivering in the first six months, there's no chance they'll deliver. You might have hired someone thinking of the next five years, but if that individual is not like able to cross the line in the first, I'd say three, six months, You know, you're getting nowhere…”

“I do believe that if you're getting someone new into the business, micromanage for the first three, six months, uh, because context is the, is the thing in a startup, right? Because your documentation is not great. Your processes are not great, et cetera, et cetera.”

“The reason why consumer-based agents are not ready to go is because it still assumes humans are watching and human will correct.”

“sales is like a mercenary profession. It's one of those jobs where you hire people and fire them every 90 days if they don't produce and there is no place to hide. Like if your software code didn't release on February first and slipped by a week, nobody gets fired, you know, because it is considered to be artistic…”

“if you're a decent AI company with a good team and good funding, chances are pretty much every large enterprise will do a free POC or maybe even do a pilot, a paid pilot. But do not think that, uh, that means you landed.”

“The simplest thing is you also don't get promoted for buying IBM. And this is the same thing applied to, you know, Salesforce and all of these companies now. They're all like safe choices.”

“I always say developers are the most disloyal users there are. And they're also constantly trying to see if, if any product tried to keep them without portability, they will also not use that.”

“nothing of values ever created until it is sold. I mean, it's such an important thing. You could be a profit or nonprofit company. I don't care. You are selling and any engineer who thinks selling is beneath them. Any marketer who thinks my job is top of the funnel and demand is completely dilutional. The point is,…”

“I think that's, it's no longer an adoption challenge. It's not, it's not an objection. Um, it just, um, there's some contradiction in, in the way these businesses operate today. That means they need to adjust to, to really get the benefit.”

“startups usually do not hire people with like MBAs and who would come from management consulting for their, in their mind, uh, you guys, you know, all about process and data and analysis, but as a startup, you do none of that, right? A startup is all about taking actions and, and, and, and learn from that experience…”

“In a long run, it's much easier, I believe, to pick a winner where you're sitting on in the back end versus on the front end, unless you are a data, you, you own the data, like Salesforce.”

“other challenge you could also happen is if you go bring a very large customer very early on, then that customer is going to force you product in a direction you may not want to go, right? I mean, if, if you bring in, I don't know, Apple as a customer, more than likely, like, Apple is going to dictate your roadmap”

“what they have sold is less important because we can teach them how to sell our product. We can teach them about the industry. Uh, so we are not like, uh, uh, only hiring A's who actually are coming from a similar space. We have A's who come from very different industries.”

“I believe in venture capital, you shouldn't make rules. Uh, in fact, our job is to break the rules.”

“the technical inflection point was AlexNet, because AlexNet is what demonstrated that by scaling a neural net, you get disproportionate outcome.”

“production companies make money and are And see just much higher adoption than developer-focused companies.”

“ServiceNow and Salesforce and Workday and all of these companies are being pushed to innovate. Like in other cases, they're like, we have large go to market motions. There's a revenue that we can easily protect and you don't have to, but because they see these drops, which are like, 20% drop in your stock price because…”

“in their top three priorities, killing Harvey will not be their top three priorities. I mean, they're like, how do I make the next hundred billion in revenue? So, I mean, that is their thought process. So then Harvey can always be like, okay, then my top goal is to build the best legal AI system. So then that's how…”

“In, in my view, where a startup can at least have a path to a hundred million in revenue, and this is still theoretical here to get played out, uh, in practice, just from observations from, from our portfolio company executing, is institutional knowledge in your domain will surpass what models can build for that…”

“everybody is now saying, let's build first instead of buying first.”

“every API request is literally that has a dollar value to it. Like an Opus request with 10,000 tokens is probably 23 cents, and this is, you're making a thousand requests, which means this is 23 dollars worth of API requests that you've made today, or two 30 dollars, whatever, that map. That is what makes this gateway…”

“You cannot put consumer as mass mass premium and premium. That's a very, uh, I would say unintelligent way of putting the consumers into that. There are only use cases you can put it that way, right?”

“all large outcomes in the world have been built from antithesis and not from thesis. So, when you question, like, this is the way things happen, people will brand you as, like, you are the niche and all of that, but when this becomes the norm, the other one will become the exception.”

“very similarly, right now in AI, we are doing LLMs, which is a different form of intelligence than human intelligence. So, uh, they have something and they are playing the sort of let's make LLMs bigger and bigger and smarter. And that is a market as well. But To build human-like AI, I think it has slightly diverged…”

“A database vendor can't capture them because no, they're not in any one database. An analytics vendor cannot capture them because it's, they capture the outcome, not the, the why and the how. Uh, and the only people who capture the how decisions are made are the orchestrators of those decisions, and you can infer the…”

“what I've learned is that it's usually cost neutral, Right? It's not like it's necessarily costing you more, because you are already living in a house, so the house has equity value, right? As a, as a house of a certain size, you are paying lawn, you know, the lawn care, the pest control, the annual fire checkout, you…”

“I feel like this is not a problem that a CRM provider can solve. This is a problem because you want to build these solutions. Every company operator is firstly unique in their tech stack, right? So it's like, it's easy to say, yes, everything integrates to each other, but there are nuances of, like, you know, somebody…”

“In the Bay Area, the cost of that talent, you know, is very expensive. Whereas this talent is in Toronto, which has a great University of Toronto where all the AI researchers coming. The cost in Toronto is low. The churn is less.”

“no matter how hard you try, if you have picked the wrong market, if you have built the wrong product, There's no amount of hard work that's going to change the fact that the market doesn't want what you have.”

“great founders don't build markets. They find markets that exist and then they, they quickly, uh, raid those markets.”

“And I, I think like this concept of Always, always pushing or squeezing the last bit out of a product, a technology, an opportunity, or a team, um, is a hallmark of the highest performing teams and companies.”

“if the answer is what we met in the incubator six months ago, And we decided to start a company together. I'm like, mmm, like yellow flag because you guys don't have high trust yet. You haven't spent enough time living life together. You don't have shorthands and having spent two, five, eight, whatever the number is…”

“your odds are improved When you spend time seeing high growth companies from the inside, seeing them go from zero to a hundred, zero to a thousand, whatever metric you want, but watching the chaos and the insanity and the intensity that exists only in hyper growth companies, you have to see that from the inside. Once…”

“don't do the income statement if there is work to do over here on the product, because you can always clean the income statement up later, but the minute you, you leave a crack for a competitor to come in and steal your lunch, your lunch is gone. And you just have to have that mindset. And I, and I do think it's like…”

“when you ask Anybody who's good, would you have me make it easier for you? Or would you have me give you more work to do so that it's uncomfortable? The highest performers, 100% of the time say, I would have you give me more work. I would have you load me up. I would have you push me to my limits. That's why I'm here.…”

“Founders should be maniacal about some arbitrary thing in the product. And never, ever let go of it, because it gives them the perspective on the system that they'll lose the day they give it up.”

“So, so I, I describe your Delaware Interperation as darts drift to Indian founders. You know, if your customers are in the U.S., if your investors are in the U.S., you're much better off interpreting Delaware.”

“If you are hung up on the valuation in the bid need and you're spending six months a year to raise money, you're being foolish. You're being left behind. You could double, triple value if you took the money and started to build this stuff.”

“So, uh, you see, you, the secret of early stage investing is actually to invest behind founders and teams and companies that are going to succeed anyway. Maybe we made two percent difference. But you've got to be humble enough, enough to recognize that.”

“And design is such a thing, it's very hard to offshore, because it's so cultural. Just imagine if, you know, someone has not experienced Disney, and they have not experienced how healthcare system works, or they have never bought an automobile insurance in their life, they can never connect with these companies.”

“We talk about mentorship. We talk about networking. All of that is, it, it is important. It is good. But nothing as vital as sponsorship. And ICICI banks sponsored women for the top jobs without all these other programs, you know, and I think that is what made the difference.”

“Failures have high anecdotal value, but success teaches you like no one else.”

“If the founders or none of, if the founder or the co-founder is not even close to the innovation circle, innovation area, I feel like they may be cut out from what is going on.”

“many of these new AI enabled businesses that you're building, whether it's pure software, whether it's services, Has a Palantir-esque, ah, ah, flavor to it. That, that in delivering this service and in delivering this software, ah, AI enabled software, you will need this, you know, the way Palantir built it, what they…”

“If I'm going and getting design partners in my network, because they, there's some prior relationship, they are biased to give me feedback that I like to hear.”

“even if a founder has raised multiple early rounds, Sid, ah, the founder, the exec team, and the larger team should never be paid the market salary.”

“I would say don't bring humility into storytelling. Be unapologetic about it. Present that grand vision and demonstrate passion.”

“Incumbents have the advantage of a lot of data, right? But they also have the baggage that often the data is not very clean. So as you build your models, as you build your foundation models, et cetera, are you relying on good data or not, right? Whereas a startup has a clean slate, right? And they don't have that…”

“Then you realize retention curves don't matter, finally you need to make money. You start building business models on top of the retention curves, then you realize the retention curve will drop the moment you put money on top of it.”

“So if you don't generate demand, and if you say, I'll give you a tool, it's an operational improvement tool. Most of the Indian businesses are anyways very tightly run. They're not having large man for, ah, manpower that they will need to cut people. One person with two people are running a shop. Usually a single…”

“Hardly any good deal has happened where there was complete consensus. So, because outlier outcomes will only come when there's no consensus. So one person has to truly, truly, strongly believe in it, and then of course debate, fight it out, get other people feedback, but it has to be one or two people liking it much…”

“comparing with US is wrong, because US is 15 years ahead of us. And when you, when you have one hundred billion dollar company, Uh, being created, then for the next 10 years, that firm is, is very, very aggressive.”

“Real deals only happen after continuous engagement because that's when the founder also knows you. That's the only time. Otherwise it's a FOMO investment.”

“If you're not going to be surprising competition, which means there's no advantage, ok? Forget it.”

“So I always say rich people buy brands for vanity to show off. Poor people buy brands for security.”

“Don't get swayed by a CEO who talks rubbish, who makes grand plans, five years you'll be here, 10 years, don't get carried away by that. Because a lot of CEOs promise and don't do anything.”

“in most MNCs, a budget setting exercise is an exercise in minimalization. What is the least I can get away with?”

“So, always look for what I call the full number manager, not the decimal point manager.”

“You know, a lot of founders, the other mistake they make is, the first sign of rough weather or turbulence, they cut price. In fact, I would say the first sign of turbulence, invest in innovation. Get something new to the table. Because in every troubled times, If you do the same thing again, the comparative benchmark…”

“A lot of CEOs outsource their presentation to a junior manager. Don't do that. It's your voice. When somebody else does it, it's not your voice. It's a voice over. A voice over is very different from a voice.”

“In these industries, product mattered, innovation mattered. In FMCG, innovation didn't matter. They all talk about it, but there's no innovation.”

“Agentifying workflows are getting very simple. So the workflow is not, not a mode. Being embedded in a workflow is not a mode because switching costs are getting lower.”

“design is such a thing. It's very hard to offshore because it's so cultural. Just imagine if, you know, someone has not experienced Disney, and they have not experienced how healthcare system works, or they have never bought an automobile insurance in their life, they can never connect with these companies.”

“If you are able to get Indian companies pay a hundred K plus and even half a million dollar plus, um, kind of in terms of ACVs. I mean, it's, it's a big, I would say validation of the product market fit.”

“people always say that, hey, if you haven't built the RFP, you haven't helped to build the RFP, it's hard to win. But we have actually had a good experience. Uh, so we were able to win in multiple RFPs, which we did not help build. I mean, obviously there are also RFPs that we might have lost because we were just in…”

“You can, if you've got the correct processes in place, you can actually get a half decent engineer and train them on your processes because AI is very great at even reviewing your entire code base and infrastructure and giving context to a new person and make them productive very quickly.”

“Google sees that Astra as a brand is putting out a lot of content on YouTube or other channels, and other people are actually putting a lot of content about Astra and marking it as a top-end test company, and it has nothing to do with backlinks. Then Google will automatically know in an air first world that this is a…”

“My challenge to a lot of folks building in this BtoB SaaS space is try putting that five, 10,000 dollar plan on your website, and you'll be surprised Uh, how many people, especially in the West when doing B to B buy are comfortable swiping their credit card for that 10,000 dollars because that probably isn't that huge,…”

“certainly I think the value of experience Has diminished in the value of intensity and curiosity has exploded.”

“market and timing is, uh, probably very important, much more important. Then, um, how good of an entrepreneur you are. I think, uh, uh, you might be the best entrepreneur in the country, but if you are working in a small time or, uh, if you're starting five years too early, this is nothing you can do about it.”

“I think there's the second myth that is there, the Indian customers Want endless customization. That's not the case. I think customers are also realizing the power of platform. We have built this like a platform. We've made sure that the platform is configurable to cater to different use cases. And that's what Indian…”

“SaaS is becoming more of an execution layer, not an intelligent layer. Now the intelligent layer is moving more towards, uh, the, um, AI agents, where it is more autonomous in nature. Where, ah, where, when, when employees interact with it, they don't interact with the SaaS application, they'll interact with the AI…”

“diversity of experiences is very valuable. So if everyone comes with a different history of what they know, then obviously as a team, you know, a lot more than what you would know individually, but a diversity of value system, uh, very early on in startup, uh, is, is, um, not important, uh, not, not important. I mean,…”

“Diversity value systems are great for democracy if you're running a country, but if you're running a startup like a democracy, you'll never end up making any decision with a fast enough speed that a startup requires.”

“So what not to build is really to not to build the, um, Things that you feel very passionate about, unless you're very, very super sure that that passion is shared by a wider community.”

“So right thing would be to hire someone in the outbound, even if it's an outbound STR, but probably report that outbound STR directly to you as a founder, as a CEO. Because new functions do require that, um, hundred percent attention and focus for them to initially bloom enough to show any results.”

“Startups are perfect for commercializing what knowledge exists, and that, that's what it should do. But fundamental research, of course, companies like SpaceX exist. But, uh, I don't think startups should pursue that.”

“Uncertainty is actually a moat for startups. Um, and I would differentiate uncertainty from risk. Risk is something that you can model on an Excel file. Risk is what you can say. I have a five percent chance of failure and so on. And if you're able to quantify risk, then big companies are able to pursue projects.”

“Transformer models, OpenAI, is great at providing feedback on the average, but it doesn't cover the edge cases. Or the edge cases of your opinions. So that's the problem. So ChatGPT is good at persona feedback. It kind of averages it across the population. But if you really want to understand how does the population…”

“I always say that, uh, when you, uh, a solution coming out of India will be completely different for, for the same problem. Let's say you give a problem to, to, you know, to, uh, Nvidia and to a local company here, right? The solution will be different, right? Um, And I think our solution would be a lot more practical…”

“sometimes I feel it's much more important to, I think trade some amount for relationships over there, because it's not the only time when you're going to work with each other, but, but I think, uh, there will be multiple occasions where you're going to work with each other.”

“So I would, I often say when I meet entrepreneurs, I, I think look at the market size with a little bit of skepticism. And typically in India, what I've seen is the market size is smaller than what you think it is.”

“To also kind of assess during those early conversations, where are they in their learning curve? Are they at that denial point or are they at that point where they say, look, I know this works. I have tried my hand myself, but I think it's best done by somebody who's doing it as an institution. Only then you can…”

“the number one thing for founders, right? No matter what your brand is, how much value you have, whatever you do, the valuation matters, what price you're giving them money at, how much ownership you're taking. That is the first priority for them. Everything else comes second, third, fourth, right?”

“I think the, the, the, the way to conclude it is that, you know, the engineer should become good salespeople and the salespeople should acquire the skills of engineers.”

“The criteria is that if there are other human beings involved in the actions, I mean, consequences of an action, then you would need a human being.”

“I think we underestimate the talent that we have. We overestimate the talent that we want to bring. Um, the board always thinks this people, this is the biggest job this person is doing. Uh, tells the founder that go out and look at people. Uh, I think consistently, uh, I felt, uh, the talent that has seen ups and…”

“Second thing is also they will be very candid with founders because they have been there for early days. If you do something stupid, they will tell you it's stupid, but most new executives, they, Tell your founder what an amazing company you built, how great you are, how thoughtful you are, how smart you are.”

“Actually, IPOs do most collateral damage to companies, you know, because you bring a lot of CXOs. No, actually, not don't take company public, I think, be confident to take a company public, and while retaining a lot of your founding leadership. With your existing leadership, rather than bringing, at least, like, if…”

“it takes a year for anybody to figure out the company, a year to actually take bets, and year three to really deliver value”

“when you're doing, let's say, one, two, three, three, two, nine, that initial journey, I think, uh, generally, 99% of the times, I think your time is not a problem, because I think you have enough, uh, headroom, enough, uh, room to grow. It doesn't matter how the industry is performing, how the macros also relatively…”

“Indian customers are, uh, you know, not super easy to please, uh, for sure, but there is, uh, they're also tolerant. I mean, I think it's, uh, it's not entirely, you know, true that, uh, they're just one way to look about, uh, the customers. I think, uh, um, they're willing to partner with you to build use cases and…”

“the way we've designed some functionality in our product, which is relevant for high velocity sales, especially when you're selling to consumers. Um, so when you're used to that process, uh, it'll be difficult for you to then go back to a B to B type of system and try to do the same thing there because you lose…”

“Now what's happening because of AI is one more business model innovation or change, not software as a service, but service as a software. That is the latest thing that's happening.”

“And from the point of view of large FMCG companies, This is one of the way to create. They cannot create zero to one. You already have created a simple zero to one. They can take it and actually go for distribution.”

“And at least what I have observed is a company which you give, let's say a half a million versus a company which you give a five million. Sometimes the half a million company scales much faster, better than the five million company, because the half a million company has only so much money. So he has to find the kernel…”

“It's okay if you're a rapper on GPT, because you want to establish customer value first. If you have established customer, if you are getting customers, everything else will follow, right?”

“I'd say, 2017 to 20, 21. A lot of these companies that were, were compounded companies in terms of their architecture, ended up thinking they were Silicon Valley companies and raising whatever, hundred million dollars, five hundred million dollars, whatever that number is. But at the end of the day, gravity sort of…”

“A lot of these companies, if you look at these ones, they're hybrid India-US, Where the founders have moved to the US and tapped into that market. I think it's hard to build a software company that's gonna win in the US where the founders are only in India.”

“So what we've seen at least is in this category, companies start off with 30 to 40% gross margins, because they're doing a lot of custom work, and they're figuring out what the product is, but once the product gets more productized, and there's always human in the loop, but it gets more productized, you can get, you…”

“if you look at venture capital, if you look at most businesses, almost everything can be copied in no time. Reputation cannot. Reputation is a moat. Reputation compounds with time.”

“When you look at Tata, Ambani, Birla, Adani, all of these guys, they are in multiple businesses. That's considered a feature. When a tech entrepreneur wants to do multiple businesses, it's considered a bug.”

“Bhavish made his business profitable before he went and did something else. You can't be loss making in one business and start loss making another business, right?”

“Every time we have violated that principle, it's our anti-portfolio. Every time we have followed that principle, it's a great hit in our portfolio. So Ola, we didn't believe in the market. We believed in the founder. So we have a line internally when somebody says, like the founder, don't like the business. That means…”

“The fact that we could avoid monopolization and wrong kind of toll collecting, uh, because the primitives in our view should be public utilities, uh, data blind public utilities rather than anything else.”

“We get a little bit of a chance. We go from being the underdog to the top dog, and we think everything is sorted out. And this is our problem even, even in DPI. And, uh, uh, so it's very, very important that we maintain an underdog mindset.”

“GST revenues are very hard to fudge because we do, We, we do the value chain, uh, uh, kind of settlement, not on a monthly basis, not even on an invoice basis, but on a line item of an invoice basis.”

“But what will happen is that if these companies are profit maximizing entities, they will suck the value away from the edges, because if you, on the other hand, make them a public utility, Then the value is either with the person who had the data, the hospital, or with the person who made meaning of that data, which is…”

“Our view in India is that we don't know how the ecosystem will evolve. There's a lot of unpredictability. Therefore, we must think of it as a complex adaptive system and retain the right to intervene wherever it makes sense. For some years, it may make sense to do it at the app level. Some years, it may make sense to…”

“So we have to take some agency away from the children and give it to the parents, because if you were in the West, we would give it to the teachers. If you were in China, we'd give it to the state, you know, but in India, we have to give it to the parents.”