Full Transcript

Dhruv Sharma: Hi there listeners. It's March 9th. We're streaming live. We've got a lineup of three very exciting guests for you today. We're covering ad tech in the age of AI. We're talking to someone who's innovating in the home rental market, and then someone who's making it easy and fun for you to take healthy meals to school. What's up? How are you doing?

Utsav Somani: I saw the India World Cup victory yesterday. That was fun.

Dhruv Sharma: Oh wow. Congratulations to us. We won against New Zealand, right?

Utsav Somani: I was excited in the first 30 minutes. I think India just took it away from New Zealand pretty quickly. So congrats on winning the third World Cup to Team India. I think the entire country was celebrating yesterday. I saw some visuals from the India gate, literally, like I think everyone was out on the streets until midnight. Fireworks, dholwalas, everything. So good. I hope everyone's productive this Monday though.

Dhruv Sharma: Yeah, you ended Friday's show on this note that there was going to be a match and I'm so glad we came back having won the match.

Utsav Somani: Yeah. All right. So let's get started. David and Daniel, welcome to the show. Thanks for having us.

Utsav Somani: So let's, I mean, we're a quickfire format. We quickly just jump into the question. So anytime you feel that we're going too fast, like feel free to like just take a breather. So every ad deserves to have this creative agency. That's Loud Echo's thesis. Can you explain what that means?

Daniel Keyes (Loud Echo): Yeah, absolutely. So David and I are brothers. We co-founded this company. We've been building together for about a decade. And what we built with Loud Echo is an entirely new way to run ads online. And what we've done is we've taken the kind of traditional advertising platform and we've rebuilt it from the ground up and embedded real-time generative models at the core. And what that means is today we can understand every context, every opportunity that every ad serves and can actually in real time generate a better performing ad specific for that audience and context and kind of impression level details. If you think back on the last, you know, 20 years of marketing tech and ad tech, almost all of the machine learning and tech has focused on who to, you know, target, how to reach your customers and then how much to bid. But the actual creative, you know, arguably the most important part of a brand telling its story to audience stays entirely manual, entirely separated from the buying and optimization platform. It's literally two different silos today, right? You go to Canva, you go to Photoshop or others, you design an ad and then you kind of copy and paste it over to one of the most automated and optimized platforms in the world. You know, ad tech is incredibly measured and optimized, but the creative itself has stayed outside of that loop. And in the age of generative AI, in the age of endless content, the ability to tell a million stories to a million people, we're empowering brands to be able to kind of take back control and tell a unique story to every user. So that's the 30,000 foot view.

Utsav Somani: And you do this under 300 milliseconds, like bid, generate and serve. Like what's behind the scenes? Like what's the technical stack? Like what image generation models do you use?

Daniel Keyes (Loud Echo): Yes, we use a combination of models. It's kind of like an orchestration engine, if you will. Some of the things we're doing in the bid window, in real time, kind of add literally as the page loads. And then we have optimization and caching and kind of a whole skew of other techniques that we use. And so the ad tech is this beautiful place where you get, you know, Internet scale decisioning that's happening. And that's happening today on brand safety and contextualization and targeting. But it's not happening on the creative decision itself. So that's kind of what we're pioneering. And, you know, we're kind of riding the wave of all the great innovation on open sourcing models and making them faster and smaller. And, you know, this trend is just going to continue. And we really believe that the future of advertising, the future of brands telling stories is going to happen just in time. Right. It's far more effective to ask a model in real time how you should be communicating with someone as opposed to pulling an old version from an A-B test that was started weeks ago and doing kind of some little A-B testing on that. And so that's definitely the direction we're going. There's still way more optimization to do and we can speed it up a great deal. But so far we've seen, you know, 3X improvement compared to static variants.

David Keyes (Loud Echo): I would just add for context that it's the most exciting time to be in this space because Daniel and I for the last 15 years have been tinkering with this idea, you know, as brothers walking and having great conversations. How do you get the right message to the right person at the right time? And, you know, Daniel was in the tech world. He was in this elite R&D unit and I was the prime minister spokesman. And we would always ask the question, you know, how do you get something in front of people that they love, not just they tolerate? And ads for a very long time were just subpar. Anybody who knows this who's been online, it's just we don't need to be suffering with these crappy, terrible, horrible ads. And so the tech that we're building is really hoping to deliver the best message at the right time to the right people.

Dhruv Sharma: You guys are making ads cool again, otherwise ads would become like this bitter pill that you had to swallow. But I want to understand a couple of things, even architecturally. So it appears to me that the latency window was, give or take about 300 milliseconds in the cookie and auction model. It's still 300 milliseconds. And then you've added generation, like creative asset generation on top of that. Is it a fair assumption that you're optimizing other things that also take place simply so that you can create a time window for your, you know, for your models to generate the creative asset?

Daniel Keyes (Loud Echo): Yes, so there's kind of the the restrictions of the real time system that exists just to bait and target and serve. Right. And we kind of we stand true within those frameworks. Otherwise, none of it works. But what we've added is the ability to kind of run an online learning process across your creative. Today, most of the optimization happens on the targeting and the bidding and kind of maybe a few selections. Right. The kind of gold standard is A-B testing or dynamic content optimization where you're choosing between a number of variants. And what we've done is we've built kind of a reinforcement learning framework that now can add the creative aspects to that. So some of the things happen offline, some of the things happen online and in real time. But the underlying impact of that is that now you can you can serve orders of magnitude, more versions, and you can make decisioning kind of in in real time as to what what someone should see and how you should communicate with them. But it's it's really interesting to see kind of this technology get integrated back into the into the creative side because it's it's something that that's so desperately needed. Right. Like, you know, it it makes no sense that we create real time decisioning on who to target. But the narrative, the the the the creative aspect of the campaign is decided once and then, you know, tens of millions of impressions then see that. So that's kind of what we're trying to to effectuate here.

Utsav Somani: And what kind of industries are you working with right now? I believe you're working with ecommerce, gaming and a bunch of others. So what kind of data points go into making these ads smarter, like in terms of every impression acting like its own agency of sorts? Right. What are the data points that you're tracking return on ad spend maybe for some industries? Some are just pure impressions. So what kind of data points go into making this ad or the creative more successful?

Daniel Keyes (Loud Echo): Yeah, so it's true today we work across pretty much every vertical, you know, finance, fintech, real estate, e-commerce, DTC, really any company that's already spending a significant amount on their advertising budget, if it's the part of their growth strategy and they need to reach a new audience and tell unique stories, we're we're a great platform for them. And as far as measurement goes, you know, some of the measurement that exists in the industry is wonderful. Right. There's there's a bunch of we're not reinventing the wheel in all aspects here. So there's a lot of audience related measurements. There's, you know, how engaged you were with the thing that you're watching. There's clicks, there's conversions. There's very far down the funnel things. And essentially what we're doing is we're taking all of that and then we're embedding it into the kind of training loop of, you know, what what is the story that we should tell? What is the end state here is that there's a model, there's tokens that that fly in in real time. And then and then you just get the ad on the other side. You get the story. And where that becomes really helpful is, you know, today, basically brands have to tell the same story to everyone. And in the platform that we're building, we envision a world where you can tell a million unique user journeys. Right. So the first time you engage with the brand and the 10th time should be different. It should be a different angle. It should be kind of a fresh story. It should keep you surprised. The breast, the sorry, the best brands do this already, but it's impossible to scale. And if you think about kind of banner blindness or ad fatigue, it's an all time high. Most people just don't even know what they're seeing at any given moment. And the explosion of content is really kind of responsible for this. And what brands do is they cap the amount of times that you can see their campaign, which is not the most optimal way to do it, because if you can tell that someone, you know, would like to hear from you, then you should be able to tell them that story. You should be able to communicate with them. And so what we do is we allow brands to kind of tell a unique angle, a unique story every single time that there's an impression from that from that audience or from that cohort.

Utsav Somani: You work across all platforms, if X, LinkedIn, Twitter, sorry, Instagram, Facebook, all of these platforms are compatible with everyone.

Daniel Keyes (Loud Echo): So today we we focus on what's called like the open web. So we serve on mobile apps and games and publishers kind of anywhere that we can. By programmatically through, those are the places where we've rebuilt the entire stack, so today the platform kind of integrates the targeting, the buying, the bidding, then the the ad creation, the delivery and the measurement, and we run that flywheel kind of over and over impression to impression. And we're working on pushing into the kind of quote unquote walled gardens, the tick tocks, meadows, et cetera, of the world. What's going to happen, I think, in ad tech is that today, the. The kind of command control and command center is post campaign, it's post click, you go to places like Sprinkler or others that have kind of all of your data from all your campaigns, the the social platforms, you know, and kind of every other channel owned and operated that you do, those are the places that tell you over all of your budget what's happening. But those are after the fact. Right. And those don't take into account all of the targeting capabilities and all the different variants of the creative. And so what's going to happen in the age of generative is that that's going to go to pre-click, it's going to go before the campaign. And we think that a platform like ours is going to be the perfect place to kind of understand how to communicate to your audience wherever they are, because now kind of with our system, you can engineer a campaign that talks uniquely to every customer that meets them at the journey point that they are. And then, you know, you click a button and that kind of happens autonomously. So it's a really big shift that's coming. And it only comes online once you take seriously this world of an infinite amount of content. It kind of breaks, breaks the paradigm.

David Keyes (Loud Echo): And by the way, you asked about the brands you work with. Most are in America, some are in Israel. Part of the reason we were excited to speak to you guys and Daniel and I trekked out to the huge AI conference where we met you was to expand our vistas. And India is a growing and massive and amazing market. We met with some of the big brands, some of the big agencies. There's a lot of really interesting stuff not only happening in AI, obviously, but happening in the advertising space in India as well. So we're trying to reach the world and anybody who wants to reach the world can use our Loud Echo system to do it more dynamically in real time. And more effectively, most important.

Utsav Somani: That's amazing. And so glad we met at the India AI Week as well. India is extremely price sensitive. Does your platform give price advantage as well in terms of just being more effective at bidding, saving costs there? Of course, generative AI is being used to create the creative. So that must be cheaper than a designer as well.

Daniel Keyes (Loud Echo): Yeah, over the long term, yes. I mean, there's definitely kind of a learning curve. There's a kind of a higher CPM depending on what type of campaign you're running, how many variants and kind of real time creative you want to be generated. But yes, if you if you snapshot us now, if we speak again in a year, it will be not only far more effective, but far cheaper to to run campaigns like this.

Dhruv Sharma: So even now, Daniel, even now, does the cost of generation make sense, given, you know, the the existing CPM waterfall? Well, you might imagine it would get lower over time, but.

Daniel Keyes (Loud Echo): Yeah, so so I was just going to touch the India CPMs and kind of America CPMs, they live in a different kind of universe, so there are places where it could be a little bit price prohibitive right now. We're working on ways of kind of integrating this technology no matter what the kind of inventory and buying market is. But but yes, the ultimately, you know, the cost of the media is is, I think, downstream from kind of how vital or what opportunity lies in in in that in that impression. And and then even more so, the kind of system becomes important because it allows you to take advantage of that. Right. There's there's kind of there's there's firms that do, you know, algo trading on on kind of the ad tech market and they try to optimize the bid. But when you already know that you're going to show someone a message, what that message is becomes even more important, especially when you have more and more inventory.

Dhruv Sharma: You know, it's very interesting you bring that up, because when I read about you guys or first heard about you guys, I think the first thing that occurred to me was this is so much like high frequency trading. I mean, maybe you're talking milliseconds here. They're talking microseconds, but it's kind of similar. You get the you get the bid request and they get certain signals about the orders. But one of the questions. So so lots of us about price sensitivity. Let's assume for a moment that Indian ads and Indian advertisers also very conservative in terms of how their how their brand gets perceived. And maybe there's a I'm going to ask you a question. Maybe there's a lesson in here for Indian authors as well in terms of just closing B2B sales. How do you convince advertisers that whatever is going to be generated runtime in terms of the creative asset will be on brand? It's not going to come out of a pre-approved brand catalog, but it will still be on brand.

Daniel Keyes (Loud Echo): Yeah, it's it's it's the most important question, and it's kind of been central to everything we're building. I think brand safety is a non-negotiable. And I think us as entrepreneurs broadly, our role is to try to see things a little bit before they happen, to see where the puck is going. And that's kind of how we've tried to engineer the system. So we've built a bunch of protocols and kind of brand safety guardrails. Throughout the process. So before we buy an ad, we check the inventory. We check the kind of contextual safety of it. And then as the creative is being generated, it goes through a number of you can think of this kind of like deterministic agents. These are kind of queries that are per client, per campaign that are being run in real time against an ad. And then we have a kind of a sophisticated routing system where if something is flagged, we serve a kind of default pre-approved ad. And then, as I mentioned, we have optimizations that happen kind of across the arc. They're kind of near real time that help push forward a lot of improvements as well. But there's really a number of trends that I'm seeing that assure me that this is where we're going. Right. So one is today already you have non-deterministic agents running in the loose, doing real work for real companies in the form of customer service agents. Right. There are, you know, an uncountable amount of them that are answering phones, sending emails, engaging with real clients with real money on the line. And those companies have built systems around them that give the kind of managers at those companies the ability to kind of operate them effectively. So that's one. It's already happening in the live at scale. The second is, I think we're all kind of noticing how these machines of thought are becoming better and better. They're kind of adhering to our guardrails more and more. And what we've done is we've done kind of things that don't scale in outside of the industry, but applied them very specifically to our use case so that we can see the kind of deterministic nature that we need to from the brand safety side. You know, but then the third, I think, sorry, go ahead, Dan, and then I'll just finish the last point. The third and I think most important is if you read back the history of ad tech, you know, marketeers used to decide exactly where their ad showed up. Right. They would go to that publisher, to that page on Popular Science, you know, what I would read or exactly the banner on the article. And they would say, this is where our ad is going to be. And then inventory exploded and Google and others launched kind of the programmatic age where you could buy and trade and place and brands had no idea where they were going to show up. This was a huge issue, but it worked much better. And the tools were built in the industry to make sure that it was safe and to have kind of remittance or anything else that needed to happen as a result of it. And the same exact thing is going to happen here. We're going to learn to trust these agents in real time. And we built a system around that. So there's human auditing, there's mitigation, there's an entire platform that the brands get access to that helps them kind of control everything. And it's hugely important. We wouldn't be able to run without it.

David Keyes (Loud Echo): Well, we have about 30 Israeli, mostly engineers on our team that are building these safety mechanisms. But I think like with any great disruption, there's going to be early adopters, there's going to be people who fear it. But the writing's on the wall. The era of Mad Men is over. It's not a few people sitting in a room deciding what everyone in the world is going to see. That's gone and done with. And so, yes, there will be some industries that will say until I see every single ad, nothing is going to go out. But that's going to last a very, very short while because the results are going to be infinitely better and the process is going to be cheaper and smoother. And frankly, the safety mechanisms will be so good. It's not just it's not just AI running. It's also AI safety measures that we're working very hard to build. So the writing's on the wall and we're trying to build it first and early. And it might not be for everyone today, but in six months or in a year, everything will change, I believe.

Utsav Somani: I mean, AI is going to upend so many different industries and you're right, like the era of Mad Men probably is over for the advertising industry and many other industries as well. But tell us a little bit about the claim that you make that 80 percent of your qualified calls end up converting. And who's the ideal customer for you right now?

Daniel Keyes (Loud Echo): Yeah, so I think this just shows that it was surprising to us as well, like the customers are much more willing to experiment this than than one would think from the outside. I think, you know, we we benefit from the kind of CEOs, you know, jamming AI down everyone's throat. You got to try it. You got to try it. You got to try it. And then that trickles down to the CMOs, to the to the marketing buyers, et cetera. And they're more willing. And then when they try it, you know, it works. So what that means is that, you know, the brands that are already spending significant spend and have high growth needs, they want to get on board and they want to try it. To be frank, we're trying to on board as many as we can. It's we're still a small team. We're growing as quickly as we can. But it's it's been fun to see that there's kind of real, real interest on the other side.

Dhruv Sharma: And so do we have do we still have time for me? One last time for a couple of questions. Yeah, we can sneak in a few questions. So we haven't spoken about format just yet. I'm imagining image, imagining text is video on the cards already or.

Daniel Keyes (Loud Echo): Yeah, it's like you're inside our stand ups. Yeah, today we do text, we do images. Obviously, those are the ones that we can scale the most video is we're working on it today. It's still price prohibitive. Right. The cost per minute is orders of magnitude more than it is to generate text and many more images and many more than text.

Dhruv Sharma: So I would imagine the latency windows your enemy over there, at least right now before you're heavily, heavily optimized and can quickly generate something on the fly.

Daniel Keyes (Loud Echo): Yes. Latency is an issue for that. But even if you took the kind of near real time form of campaigns, you know, it still would be too expensive to kind of actually do that at scale. And the quality of the brand safety there is not is it takes many, many, many human iterations in order to get one of these amazing kind of generated videos that you can't tell that is not one prompt. Right. That's a that's an entire system and day of kind of tweaking. So but once again, we see the trend line that that's where things are going. Right. The fundamental engineering doesn't change. It's just about how we can kind of build the scaffolding around it. Yeah. So it's it's coming.

David Keyes (Loud Echo): By the way, I just thought of another analogy of what we're dealing with in terms of safety, just to just to follow up on that, you know, full self driving. Now, if you were to tell most people over 50, hey, just put your car on automatic, let it do everything. Most people probably a little afraid they might have some preconceived notions. It's not so safe. It's actually already safer. But is there any doubt in anyone's mind that the human errors will be simply eliminated with full self driving in a year or two or three? So even though there's some reticence today, it's going to be a solved problem very, very soon in the writings on the wall.

Daniel Keyes (Loud Echo): You know, I was I was speaking to someone recently who has they live in L.A., they have a young child. Sorry, maybe I don't know. They're nine. And so sometimes they need to get a ride somewhere and they think it's safer to order them a way more than it is to order them an Uber because then there's not a human driving and they trust the way more driving better and they trust it. So it's an interesting tidbit about how sometimes the scales kind of totally flip where actually, yeah, we would trust a robot driver more than we would trust a human. We don't know which isn't, you know, I think very telling.

Utsav Somani: Guys, this has been awesome. Maybe one final question before we let you go. So, Daniel, you were designing intelligence systems for the Israeli Defense Army of the Defense Force. David, you were the spokesperson for the prime minister of Israel. What got you into selling ads or making an AI platform selling ads?

David Keyes (Loud Echo): Well, you know, I mentioned it briefly, but Daniel, I have been sort of on this journey for a very long time and we were reading Kahneman and Tversky, you know, way, way back when and Thaler and trying to understand how the mind worked. And in my old job, I had to meet with world leaders every day. You know, I met with Modi many times with the prime minister. I was standing next to them in that famous picture on the water in all sorts of interesting meetings. And I really thought long and deep about how do you understand someone's worldview? How do you meet the listener where they are? How do you understand what animates them, what their hopes and dreams are? And so I was doing that in a very analog fashion by diving deep on the world leaders that we were meeting. Daniel comes from a different background. And so, you know, I was always thinking I was always in touch with these top leaders in Silicon Valley and throughout the world. And I knew that they were on the horizon, absolute game changing, game changing technologies for what I cared about, which was the defense of the West, the defense of good ideas. And I was up against state actors that were trying to undermine Israel in a very serious way on a daily basis. So no number of small people in a room will beat the sort of algorithmic thinking of China or TikTok or some of these malicious actors on the global stage. So this was our attempt to build something in the private sector that could actually benefit industries and brands and people to more effectively convey their message on a very large scale. And it's something that after the advent of LLM sort of were born really became possible for the first time on the scale that we always imagined it. But I'll let Daniel answer from his end.

Daniel Keyes (Loud Echo): Yeah, from my perspective, I mean, I did physics research, right, so I like to think in kind of fundamental systems and to me, you know, human decisioning, the human mind and the kind of creative space is another space that can be modeled. And what I saw was kind of very crude things, right? Like, you know, there's a moment of ingenuity that gets turned water down, water down, water down, eventually turned into something that, you know, a billion impressions see. And I think now in the age of kind of unlimited generative AI, we have the chance to kind of raise the abstraction layer of how we deal with marketing. So, you know, we feel like we empower marketers to be more powerful, to tell more stories, to kind of play with the state space, if you will, of creative ideas. Instead of just being one thing, it can be this kind of cloud of options around it. And that really excites me, kind of modeling states that are not well defined. And yeah, and also we like building things that have an impact on people. So it's been a long and windy road, but we're glad to be here.

Utsav Somani: All right. I hope you get a first set of Indian startups working with you. Indian tech companies as well, working with you very soon. Thank you so much for coming on the show. Stay safe. Thank you, guys. Thank you. Thank you. All right, listeners, we're moving to the segment two of our show today. We have Rishabh of Flint joining us. Rishabh, welcome to the show.

Rishabh Agnihotri (Flent): Thanks, Utsav, for having me here. Thanks, Utsav. Nice to meet you guys online. Great to meet you.

Utsav Somani: Let's introduce Flint to our audience. And what's the story behind the name as well?

Rishabh Agnihotri (Flent): Oh, that's, that's pretty interesting story. I mean, like, you know, honestly, it was just a bunch of dots on the walls, we were sort of trying to think of a name, but then... Was it like a domain name search led to the company name kind of story? I mean, you know, first you try the dots, first you try the names and then like, you know, try the domain names, something which like, I mean, what we're trying to do is we're trying to come up with random names around renting or something around flipping. And, you know, we came up with Flint, which is like a flip on renting entirely. And we looked up the domain and it was available. So the domain search wasn't a problem. The name hitting, I think it took us like hardly two days to get at. My co-founder Mayank sort of came up with it, which was really cool.

Dhruv Sharma: Do you remember what was the closest second contender to Flint?

Rishabh Agnihotri (Flent): Would you honestly know? No, no, nothing. Because you know, Flint hit us so hard.

Utsav Somani: It was so perfect.

Rishabh Agnihotri (Flent): Yeah, it was just perfect. Yes. Yes. Yeah.

Utsav Somani: I wonder if Dipinder said that when he heard Zomato for the first time.

Rishabh Agnihotri (Flent): I mean, it's a good story to have, good story to have.

Rishabh Agnihotri (Flent): Yeah. Cool. Yeah. So, I mean, the way I put it is we're trying to sort of create a new standard of rental homes in India. You know, every category in consumer has changed from like, you know, grocery to food, to everything. But if you look at the basics of India, roti, kapda, makan, makan is like really behind, you know, doing really interesting things in the market. But homes in India are still stuck in the nineties. We're trying to sort of change that by building a new standard of homes. What we do is, you know, we create these designer turnkey homes, which are move-in ready for you. The way you put it is like, it respects your taste and time because, you know, as an affluent renter here in Bangalore or in Bombay, you know, the hardest thing to do is not find a home, but is to find a tasteful home, which just like, you know, looks perfect, which you can walk up, you know, you can walk into after a busy evening, right. And you can have, you know, like a good coffee in the morning. So, I mean, we designed those spaces here in Bangalore. We make them, you know, like we list them on a platform called Fendt.in. So we're trying to solve everything from the discovery layer to the living layer of the product, which means like, you know, we help you discover great homes and once you move in, we also give you a great living experience where, you know, everything, every problem of yours, like, you know, it could be a plumbing problem. It could be a furniture problem. All of that is covered by us. So it's pretty much like a designer's rental home network of sorts.

Dhruv Sharma: Do you guys look out for like brand new homes where the, where the homeowners is essentially giving you a shell or do you sometimes even remodel like lived in homes?

Rishabh Agnihotri (Flent): So, I mean, that's where, that's what a magic place in, you know, all the new inventory in the market is generally far away, right? I mean, if you look at, but you're not getting a new inventory here. So we, we take up old homes and we flip them. That's where like Flint as a flip comes in. We have like a five-day framework where we, you know, design those homes from, from the paint of the wall to the fabric of the sofa is custom designed by us. So it all sort of comes and together we have an internal design team. So it's very much focused on design that, that way. And you know, that's completely full stack. You know, so we flip the home entirely from a bare shell to, you know, adding so far designing the paint, even designing little frames on the wall. So, yeah, I mean, we do all of that. I mean, all the maps and I think we removed that feature on our website, but, but we had this feature on all about the state where you could actually see a bare shell home and you could see a Flint home, but practically not believe that. Okay. This, this, this really happened. Uh, but we do have a video out on our YouTube channel, like, you know, describing that process quite loudly, the same person.

Utsav Somani: I mean, the Indian rental market is a war zone. Like I think in Bangalore, they were taking eight months for deposit. And I mean, it's just what was that group flattened flatmates, which used to have horror stories, horror stories, you log on to X and people are just complaining about the process that they had to go through to find a home in Bangalore. So you've abstracted all that away. And I'm guessing this is not a software problem to solve. This is a pure hardcore problem to solve. So you've the numbers that I have is you've raised 25, uh, 27 crores across four rounds, three 50 rooms, 90 to 95% occupancy. And people stay with you for at least 14 months. Uh, what is that? That's so unique apart from just design of these properties.

Rishabh Agnihotri (Flent): So we're trying to sort of, uh, solve the fundamental problem of trust in this market. Um, if you look at like everything you described right now, right. I mean, still people have to search flats, uh, to flatmates through Facebook, which is like a 20 year old way of doing things, right. Um, the, the reason of all of that boils down to trust as a problem. Uh, like landlords don't trust you no matter who you are. Uh, and your tenants don't trust the landlord, like, you know, with their deposits or anything else. Um, and that's what we're trying to do. Like when we have to fundamentally disrupted the way how landlords rent their home, so we, we have some sort of a rental, uh, sort of a guaranteed product to a landlord, which we call rent protection where, you know, after an extra number of days, their rent is guaranteed and it's on us. So it's our responsibility as a tech enabled property manager to find a tenant for them. Uh, and you know, if we don't, we pay a penalty per day to our landlords, which means like, you know, once they give their homes to us, it's totally hands off. Uh, plus we, you know, invest in their home. So all the capital sort of comes from our books right now. Uh, and we flip those home entirely. So our skin in the game is equal as the landlord, you know, their home is almost as our home because, you know, we have designed it, we have investments into it. So we've sort of introduced the skin in the game on the landlord side, earning their trust on the tenant side. You know, we've made the process extremely seamless, brought a lot of transparency through a platform that you can actually go and, you know, visit in the home, that itself is a disruption in today's market. I mean, and if you want to visit a home, you'd have to like go to, uh, you know, multiple property aggregator platforms, look around 5,000 listings, then like, you know, call so many landlords and then one landlord will actually give you a visit during your office hours. Right. So it's extracted all of that problems into like a single product. Um, and, and, and, you know, interestingly, uh, you know, the trust problem we're also trying to solve in a very different way. We've launched something called secured recently where what we're trying to do is it's, it's just like how you pay your credit card, you know, this, uh, you can pay a rent to this app, uh, and you know, you get 1% back. So what we're also trying to do is we're trying to sort of find those responsible tenants in the market who pay on time and then sort of aggregate them into a digital trust layer. Uh, and, and, you know, their landlords pay through our app. They also get the same service. Like, you know, they get the, uh, rental guarantee on us, which means let's say if you leave tomorrow, you know, from that apartment, whenever you want to, we, it's our job to actually find a replacement for the landlord. If we don't, we pay them, you know, right up to 1.5 lakhs from our pocket. So, uh, we've like sort of, you know, designed a digital trust layer, we've designed a physical trust layer. So this is the things we're doing in the market to disrupt it. And many of your landlords and tenants, although invested in your own. Yeah, that's correct. Yeah. I mean, uh, I mean, that, that, that's the kind of signaling, you know, we're trying to do in the market. But I mean, landlords are trusting us and trusting us and putting us, putting money behind the company, right? Which is, which is quite, uh, you know, the opposite in this category. So we're really proud of that.

Dhruv Sharma: Rishabh, I'm wondering if you can tell us, uh, how, how you've scaled the business brick by brick, right? And I'm assuming there's a lot of lessons in there for other entrepreneurs who are building like hard real world businesses, you know, real world assets here, so which was the, the opening geography and how did that geographical expansion take place outwards from there?

Rishabh Agnihotri (Flent): I mean, see, honestly, it's, uh, I mean, honestly, all the agility, which we had, um, you know, when we started off, you know, in Bangalore, it was, you know, us starting in Airbnb to get passive income. I mean, you know, my, my co-founders did that just to get passive income. We were actually building like an AICRO startup, just like every other tech startup, uh, in, in Bangalore.

Dhruv Sharma: You guys started in HSR or like which?

Rishabh Agnihotri (Flent): Not really. I mean, we were in Dharmagarh, so quite close, uh, and you know, when the people started coming into these Airbnbs and they were like, you know, I don't really want to move out. Uh, you know, you can just keep taking my rent, even at super premiums. It didn't make sense to us. I mean, I think just last session, you talked about India being a price sensitive market, right? But here are people are okay. You don't take my money, but just don't be pushed out of the house sort of thing. Uh, uh, so that, that sort of clicked with us. Like, you know, it's, it's a real problem. Uh, definitely there have been like, you know, lots of failures across the market, but the consumer economy is fundamentally changing in India. And, you know, we sort of took that bet that, okay, let's launch a more, let's, let's, uh, a few more homes. And that's, that's where the journey started. We launched a few homes. It took some feedback from the market that, okay, you know, uh, these are the kinds of things I would want. These are like, they wanted transparency, like, so day to day on a website, you can actually see what, what rent you're paying to the landlord, what rent you're paying to us as a furniture rental and what you're paying to us as convenience fees. So these are the kind of experiments we did in the market versus like, you know, what, what, how are the co-living companies do? And that sort of brought us raw success on both sides. Um, and I mean, I mean, it's pretty much a journey of virginity, I would say. So we've been focused on Bangalore because that's, that's been our home ground and I mean, they're like, the demand is quite, uh, well, well done up here, which still is catching up to the demand right now. Um, so yeah, I mean, like, but, but what we have had a lot of learnings along the way, because there's entire rent, you guarantee model, right. That's self-destruction. So, you know, we have to, it's almost like underwriting a loan, uh, as a bank, because, you know, you have to calculate exact amount of risk when you're taking up a home, but okay, this, this, these are the days you will rent out the home. And if you don't, you start sort of break, you start breaking your economics, right? So we're building a lot of things internally to, you know, do that, right. We build like an entire supply intelligence internally where we understand, okay, what rents are going on and what markets and trying to sort of underwrite guarantees based on that. So, I mean, lots, lots of experiments along the way, we're still, still far, far away from being perfect, but, uh, you guys also building financial products along the way.

Dhruv Sharma: I mean, you spoke about secured, but your website also talks about loan against rental income for, for the, for the landlords. I don't know if that's live yet, but it's coming soon.

Rishabh Agnihotri (Flent): It's, it's coming soon. So see, that's, that's what, right. Like, I mean, if you look at real estate as a market, a lot of wealth is logged into that market, but it's super unorganized. So with the secured as a, as a layer of trust, right. What we're trying to do is sort of incentivize both sides, right? So we've incentivized the tenant here that one person came back and incentivize the landlord as of now with the rental protection cover, but what we're also doing is by understanding more of that data, we can unlock rental income, sorry, rental loans for the landlord, right? I mean, uh, we know that, okay, this property will always rent out. So, uh, you know, I can definitely work up to the bank and sort of, you know, give you loans on, on, on back of the property. Uh, it's a physical asset, right. Which is actually generating a yield. Uh, similarly on the tenant side, we're also sort of trying to, if more and more tenants pay on the app and be able to build that trust with the tenant, we can unlock zero security deposit for them as a product, which is like a big thing in Bangalore, like you don't have to pay eight months deposit upfront. So these are the kind of financial things that you're doing. I mean, with a background in FinTech, I've always loved FinTech and like, you know, this market is very juicy that way because there's a lot of rental flow. The ticket size is very large. Um, so there's a lot of things to do on the financial side.

Utsav Somani: Let's talk about the unit a little bit of this whole model. Uh, you signed a lease or a management contract with the landlord. We do a management contract with the land. Does that expose you to liability? What is the unit economics? What does an average person pay to you in terms of, uh, uh, rent? Like what's the median numbers? What are the average numbers for this?

Rishabh Agnihotri (Flent): Sure. So an average 10 space us around 32,000 rupees, uh, which is per room. I'm saying, right. So, uh, an average property can reach a yield of more than 90 K. We do a 25% to 30%, uh, margin on it, uh, across convenience fee and a furniture rental. So something as simple like, you know, for Lenko plus, uh, sort of no broker combined. So we, we own that 25% from there. Uh, and that's what we've been doing. I think we've been, uh, doing volumes of more than 15 to 18 crores now at this moment, uh, yearly, uh, property that you properties somewhere between like, uh, 12 to 13 months as of now. Uh, you know, this, this, this depends on property to property where you're launching. Right. I mean, in certain markets, you do have to pay a higher deposit. So the paybacks, uh, before, but all you can think when 10 to 12 months is it's our average. And the average age of a person staying at your property. As I said, it's around 14 months, but, but the company is so young. Right.

Utsav Somani: I mean, generally the kind of, uh, customer because yeah. How old are they? Because what markets are you targeting next? Because only certain micro geographies might make sense for a model like this. Right. Sure. Yeah.

Rishabh Agnihotri (Flent): So, I mean, we'd be targeting anyone who's 25 plus year old and makes minimum 12 lakhs of income, uh, which is like, you know, every, almost every person that, so, uh, of course, like, you know, as a, as a market, right, we're focused on Mike, we have a micro market to micro market strategy. So we don't do every micro market in Bangalore to even if we expand to other cities like Bombay, you know, go to town, we'll not do every market there, but I mean, renting in India, $20 billion market, right. As a founder, we have to be focused. So focus on that 1% of that pipe where, you know, it's clearly like a $7, $8 billion market with, because, you know, the potato kicks in there in the $20 billion industry. Um, and then we focused on that, uh, you know, micro market to micro market, expanding, expanding, and then we did it. Now HIPAA is also becoming really interesting for us. So, uh, yeah, I mean, uh, that's, that's pretty much the focus.

Utsav Somani: Awesome. I think that's it from my side, Dhruv, you have anything?

Dhruv Sharma: I guess one last question, maybe, uh, Rishabh, like do you guys, uh, do you think you sell the home or the lifestyle, like the amenities that come along with it? Some of these I'm assuming are in gated communities and so on. So do you have anything to say with respect to that? I think so.

Rishabh Agnihotri (Flent): We have been selling the taste here, uh, basically taste of a home, uh, you know, which makes it your own today. You know, when you actually go and rent a property, your first thought is, should I invest in this property to make it mine? Right. I mean, should I put tables? Should I put that little lamp, you know, all of that, which makes your own home, your own home. So do all of that for you. Uh, you know, in a way that your taste lunch, uh, making you feel like when you're living in your own home, uh, without a lot of that investment, a lot of hassle, uh, you know, uh, so that's it. That's, I think that's where that sweet spot is. I would say.

Utsav Somani: All right, thank you so much for coming on the show. Hope to see you in Gurgaon. Yeah. Thank you for having me here. All right, listeners. Uh, we're moving on to the final segment today. Uh, let's welcome Mahesh from Basel. Uh, Mahesh, welcome to the show.

Mahesh Muraleedharan (Basil): Thanks Utsav, thanks Dhruv, thrilled to be here.

Utsav Somani: I was reading about your story. I mean, you're doing 1 million rides a day. Uh, first employee at Uber launching that in India. Now you're selling Bento lunchboxes and getting 36 crore ARR. So let's, uh, I mean, backstory, I think interesting. So let's introduce it that way.

Mahesh Muraleedharan (Basil): Yeah. I mean, Uber was 13 years ago, so I don't really talk about it much because it's just so long ago, but, uh, long story short, I was fortunate to be Uber's first employee in India. This was after, uh, so I'm a typical engineer, MBA consultant, and then got bored of making powerpoint presentations, went on to the world of startups in 2011, and I'm glad I've not looked back, I'm still in some form or shape in a startup environment since then, and, uh, Uber came across as an opportunity back in 2013. Uh, it didn't exist anywhere in India. I was, I was clueless like most, most of us are, and we get into a startup for the first time. Didn't expect it to be the phenomenon that it eventually became and still continues to be, but I think it's sort of just built the muscle of doing things from first principles and not taking no for an answer, which eventually has led me to do what, you know, it's actually crazy, uh, me and my co-founder Harini, we quit our jobs and we are now making lunchboxes and other products for kids now. It, uh, I'll come straight from Uber to Basel. Now we basically, Harini and I are married to each other and we basically, uh, couldn't find good products that we could trust for kids when, so we have twins and when they turn four, uh, there are lots of brands and lots of destinations, if I may, for babies where you can go and get all their essentials that you would trust. Right. First try is a great example in India. And of course abroad, there are multiple brands. Of course there's a lot of destinations for adults, young adults and adults, there's a lot of destinations, but this four to 12 is like this spot where we couldn't find brands, uh, in a single destination, we actually couldn't find brands that we trusted across multiple categories, which are day-to-day use and a destination. So that's where we started digging deeper and it's still, it's still continues to surprise and confounders that there is no one single brand that parents really, really trust and kids enjoy. And it's a, it's a powerful combination to get right that the child really gets excited to use this product. And the parent still feels good about buying, you know, that product for that child, because usually there is a dissonance what the parent wants the child to do or use their child. Doesn't like it and vice versa, like chocolates, ice creams. I hate buying it for my kids, but that's what they want. Good clean food. They don't want to eat. That's pretty much where this journey started. And we are operators, my co-founder, her name, she's also from the world of startups and e-commerce more specifically. We just spoke to as many sets of parents and other stakeholders as possible. And we found that, you know, lunchboxes was a specific pain point that everyone had. And if you didn't zoom out of just lunchboxes into other day-to-day essentials, there were just so many categories that I for, you know, a brand to come and say that, like, we make really, really well-designed products that you will feel great about giving to your child as a parent. And we make it so good looking that the child actually feels like he or she has the coolest product in town. So that's pretty much was the Genesis. And this happens to be surprise, surprise at 60,000 pro market. So seems to be, I mean, there's an awful lot of stuff to do. It's just getting started.

Dhruv Sharma: Seems to me that you guys allow parents to get away with sneaking in healthy food in their kids' tiffins, because it all looks very nice. I mean, the brand is super cool.

Utsav Somani: Dhruv and I were discussing your website before. And thank you so much. Super cool design.

Mahesh Muraleedharan (Basil): I mean, we don't promise that your child will eat healthy food. That's, that's a promise I know I can never make, but we try and make it as easy as possible.

Dhruv Sharma: Location trackers, because God, I remember like, you know, when I was in school, people were forever losing their tiffin boxes and lost and found would be filled with them, but, but anyway, business-wise it doesn't make sense. I know, Dhruv, we'll get to the most serious questions.

Mahesh Muraleedharan (Basil): That's one of the most often requested feature in a water bottle. Like when I give it in the morning to my child, it has to come back.

Utsav Somani: Oh, dear Lord, Dhruv, you should go to that consumer lab, actually. And test out all the things and give ideas as well. It seems like you just revealed one of their product on the roadmap. So you have a lab as well, where you test out these products with actual parents and children.

Mahesh Muraleedharan (Basil): So it basically just started, you know, back in the day, we spoke to about 900 plus parents, maids, cooks, nannies, et cetera. I used to have these 3D printed trays of the lunchbox models that we were trying. And that's how we actually got to the winning design and a lot of the design. Ergonomics and functional design is not something that you can explain. You'll have to feel, and it's not something that you can like understand and that you feel it. And you won't get insights until you observe someone using it for the first time. So that's where it started. And we didn't have anything else to do. This was one product idea, just the two of us. So we took a few months and we did that. But how to institutionalize this so that the rest of the products that we keep launching in some form or shape still has that core consumer inciting muscle behind it was what was the genesis behind the consumer lab. So the good part is now we have customers who've actually paid and bought our products, both who like our products, both who have really critical feedback on our products. So those are like those kind of customer reviews and insights from them. The two compartment bento box that we've launched actually stemmed out of an insight from multiple moms who said, hey, your three compartment fully stainless steel is fantastic for my eight year old. But you know what? I have a four year old as well. I would like something that is lighter, that is smaller, and I'm OK if it has a plastic lid in it. So that's how that came. And we just didn't want to leave such insights to serendipity. So we just institutionalized. We have a team that just does calls on a bunch of, you know, themes every week and every month. We keep revising those to see even if there are colors, right?

Utsav Somani: This is just phenomenal advice for anyone trying to build in the noisy D2C space. Like founders and operators in the D2C space, I think they don't spend enough time. Like we see something working in the US and just blindly copying it just with fancy branding and better packaging, I think. And then it probably lasts maybe three to six months as long as you can spend on ads. But what you mentioned here, I think, is phenomenal where you get to spend time with actual customers and get feedback in ways that I've seen in ideal customer profiles like yours, where it's just growing up so fast.

Mahesh Muraleedharan (Basil): And it's also the main thing we've taken very good care is to leave our own gut feels out of the final decision making process. So I don't necessarily need to like the design that we are going live with if enough customers say that. So that's something that we are very, very clear about.

Utsav Somani: And yeah, what is one recent decision that you didn't want to take but had to take because of the market?

Mahesh Muraleedharan (Basil): Yeah, I mean, it's a very, very, I mean, we actually announced it over a podcast recently as well, like electric lunchboxes. It was very obvious, electric lunchboxes, like it has like a USB port where you just like in an office, it heats up and then like you can have like maybe not boiling hot food, but warm food, right? Like the inside was very clear. There is a market. But at the end of the day, we are building a brand for kids. We are building the brand for kids that should have existed years ago or decades ago. And our decision making framework is very specific to whether whatever products or decisions we take are adding to that core thesis of the brand. So nobody wants to give electric lunchbox for their kids, including me. I'll never want to give them anything, like no matter what certification it comes with. So we decided not to do it, right, that you can argue revenue loss, but I think it's protecting me from brand dilution because you're building something for the years to come. And we are building the iconic destination for kids is what we had in mind when we launched Basel. So staying true to that, that is one example of a decision. There are multiple, like what size of water bottle to launch, how many compartments in the backpack that we are going to launch with, how many designs, what colors, etc. There are multiple day to day decisions. And none of these are particularly like backed by a lot of data that you can say this is the right decision because consumers, especially kids, they keep changing in terms of their taste, trends change, etc., etc. But we try to do whatever best we can by, as I said, remove the personal from the process and just let consumers talk to you as much as you can. And even when they're not talking, you can observe a lot.

Dhruv Sharma: Mahesh, you know, a few years ago, the whole world awakened to the presence of microplastics in our system and historically, I mean, tiffin boxes and so on. Like a lot of plastics. Do you want to educate us a little bit about what kind of materials you use, what you've learned along the way about some things that are bad for us?

Mahesh Muraleedharan (Basil): Yeah, I mean, I think one of the things that India is waking up to a lot more than we thought in terms of velocity is this whole concept of health and fitness and like investing in health and fitness experiences has taken a giant leap forward much faster than we thought. Obviously, like, you know, I'm from South India. So we pack really, really hot food like sambar rice or whatever. And like, I can't think of putting that in a plastic lunchbox anymore. And in fact, the genesis of building out a stainless steel lunchbox came from that. Right. When we wanted to give our kids home cooked meals when they turned four for their school, we couldn't put hot food in a plastic lunchbox. They didn't like the stainless steel options that we gave them. And we didn't like the plastic ones with Peppa Pig and Mickey Mouse. So having said that, obviously, you know what is bad or what is not good for you. It's the same with your diet. It's the same with materials for your kids and especially with children. Parents always, mothers specifically and fathers as well, they have a very clear gut instinct as to what is bad for them. So we are not here to say we are the safest around. We say very clearly what we use. Having said that, we also have plastic lunchboxes, which we say use it for snacks, etc. Plastic as such is not bad. When you put boiling hot stuff in plastic, material can leach, which leads to the presence of microplastics. And if you use the same one for a really long time, like years together, that can also like add some wear and tear. We've surprisingly, like I think we've been able to sell our stainless steel lunchboxes a lot better than plastic. It's very different from what, you know, would be the normal course of action for like another player who's selling lunchboxes. Probably that's where the strength of the brand people trust us to do the right thing for them. And so when they feel that they want to get like a good, sturdy lunchbox, they come to us. Having said that, I think the whole concept of saying that I'm the most healthy, that's not who we are. We tell you this is what we use. We tell you we use healthy materials and we try to let you take the final decision as the parent, because we believe that parents know what's best for their child. Sometimes they need a little nudge, but we're not here to say that we are safe. Others are not. That's not who we are. That's probably more relevant. Also, if you're doing, say, something like a cookware brand and they're actually cooking. So also not very relevant. Also, lunchboxes. Also, we started with lunchboxes, but we're going to get on to other categories like backpacks. That's going to be the next category, along with water bottles that's coming up. Lunch bags, stationery. We are essentially getting into every day to day essential in a kid's life. So as a brand, we want to make sure that we. Our strength is basically the design DNA. So we design really, really good products in-house and we make sure that we use the best materials possible and we make it look as good as possible for the child so that he or she feel that it's the coolest product in town and we keep iterating on those designs. So we do a lot of drops. All our visual designs are copyrighted or industrial designs are patented for whatever extent possible. We try to like sort of like safeguard our IP. But at the end of the day, even if someone wants to copy, they can't copy what you stand for. So and they can't copy a consumer lab because by the time it takes someone the effort to like copy what we do, I'm already ahead in terms of how I treated the product and how I've gone to the next product.

Utsav Somani: All right, Mahesh, thank you so much for coming on the show. Wishing you all the best.

Mahesh Muraleedharan (Basil): Thank you. Thank you so much. Thank you so much, Mahesh.

Utsav Somani: All right, listeners, that's it from us. We'll see you on Wednesday at four o'clock. Have a nice week ahead. Thank you.