Full Transcript

Utsav Somani: All right, listeners, live stream number 95. Today is a special for us because we've got three offliners with us. Let's welcome our first guest today, Ankit of Colipo. Ankit, welcome to the show.

Ankit Dudhwewala (CallHippo): Thank you so much, Utav.

Utsav Somani: Ankit, always a big fan of bootstrap successes. So let's introduce Colipo and paint a picture of where the business is today.

Ankit Dudhwewala (CallHippo): So, we started about 10 years back and we have been bootstrapping for about 10 years now. Colipo is a communication platform for human and AI agents. We help companies like, we have AI tech companies, Nestle, Samsung, manage some part of the communication. Usually the communication which happens with their customers.

Utsav Somani: Interesting. And how many customers do you have and what all countries are you active in?

Ankit Dudhwewala (CallHippo): So we have about 5000 plus customers spread across UK, US, Canada, Australia, India, US and India are primarily the larger market for us.

Utsav Somani: And you started off by being easier than Twilio, where you said that you're giving businesses, I mean US or many other countries, a virtual phone number basically to access their customers. But how has that business evolved as the world of AI agents and voice agents is upon us now?

Ankit Dudhwewala (CallHippo): So we did not really compete with Twilio, we partnered with Twilio. Twilio provides telephone networks, but telephone networks are not enough. If you, so the objective for businesses is not to just have phone numbers and make calls to the US. The objective is to ensure team productivity and employee productivity, understand what their teams are doing. And that is where we come in. What Twilio gives is phone numbers. We help you understand what your teams are doing, how, when they are logging in, when they are logging out, what conversation is happening, which customer is happy, which customer is not happy. Voice AI, so we, with voice AI agent, there's a lot changing. Support, I'm sure is 100% going to move to voice AI agents. But we work in a segment where we are more focused on sales. About 90, 80, 80, 90% of my customers are customers who use it for the sales process. And there we don't see much difference with AI agents. In fact, the adoption is really low with AI agents for sales as such until now.

Utsav Somani: And why do you think that is?

Ankit Dudhwewala (CallHippo): I don't think humans will buy from agents. They want to buy from other humans. And it's going to remain same as what we are betting on.

Dhruv Sharma: Ankit, do you think you can take us on the, on the desk or the console of somebody who's making that sales call? Sure. What information do they have in front of them? How much do they rely on their training? What do they do when you ask them a question they don't have an answer to?

Ankit Dudhwewala (CallHippo): And yeah. So, I mean, I don't know if this is a good example or a bad example. But one of the companies which was using it and probably one of the most popular one, I mean, we have Nestle and Samsung as our customers, but Baidu's tend to be more popular and people used to relate with Baidu. So what will happen is Baidu's will run campaigns and they will get parents to sign up for Baidu's and the services. And then they will have agents who are in India who will call these parents up in UK, US, other, and these agents are also not just in India. They might be in Australia, Canada. And then they will call the parents up to explain what the entire product of Baidu's is and try to sell them. These are not really warm calls and these are not really cold calls. So someone has signed up, but it's not possible for them to get them to a Zoom call without doing the phone call. The phone call and the phone messaging and texting comes in between before you go to the Zoom call. And that is where Qualipo comes in. Now, and like we have other tech companies as well. Bright Champs is one of them. And I think they are pretty large. We also work with Visual Website Optimizer, VWO in SAS. I think Sparsh is also an offline member is what I know. So they use it for communicating with their customers. And now when you have 100 agents, you cannot really ask people to use their own phone. You don't get call recordings. You don't understand what is the conversation analytics around it. What were the keywords which were used? For example, one thing which we do with Qualipo is that we have a certain set of USP features and we try to ensure that whenever our team is talking to prospects, they mention those USP features and they talk about those USP features. Now that kind of analytics is not something which you can get by using the handsets and a normal phone. So all those things are where we actually come in and where we actually work.

Dhruv Sharma: And if sales is the primary use case, is it a fair assumption you guys were selling to CROs inside large organizations?

Ankit Dudhwewala (CallHippo): So about 60% of my use case is sales and about 40% is between operations and support. So primarily, it comes as sales for us and sales is our entry point. We will rather target a CRO than targeting a customer support head.

Utsav Somani: Let's talk a little bit about customer acquisition cost for you and generally the industry. I am referring to this old blog post of yours where you said that 85% of Qualipo's revenue is from outside India. And you described that US Google ad clicks at $400 in your category and CAC is $20 per customer. How does that math still work in 2026? How different are the numbers right now?

Ankit Dudhwewala (CallHippo): So we have been, our entire organization has been built around searches and optimization and performance marketing like last 10 years. In fact, at one point in time, I used to call ourselves as a digital marketing agency which had a product which was Qualipo and we did not consider ourselves as a SaaS product. And we said that what we excel in is customer acquisition. Now things have changed. Now I don't think search engine optimization is anymore there. It's vanished.

Utsav Somani: Actually, you mentioned this, I'm sorry to interrupt you, but I read a tweet today in the morning itself where it said that startups now in this world of AI is 10% product and 90% sales and marketing. So I'm guessing you believe that to be true?

Ankit Dudhwewala (CallHippo): I always believe that. I never, I don't think like we have about 5,000 plus competitors and the only reason we have been able to reach where we are is because we understand how to sell and market. I don't think any product, I think the product is only 20-30% of everything. Most of it is sales and marketing is what we feel. And yeah, with respect to CAT, we have been able to manage it until now. About 50% of our revenue used to come from search engine optimization about one and a half, two years back. Now that has changed. Search engine optimization has gone to about 10%. We are trying to find new GTMs, but have been able to sail through until now, but we are still trying to find new GTMs. Performance marketing used to give us about 20-30% of our revenue. Now that has gone to about 70-80%. So yeah, we are managing through as of now with performance marketing and trying to find new GTM for the product.

Utsav Somani: And the ARR number that you stated publicly is 8 million?

Ankit Dudhwewala (CallHippo): Yeah, we are around $8 million of it now.

Dhruv Sharma: Because what are the big trends? I'm curious, what are the big trends in telesales that you're seeing worldwide? Because you have the benefit of being across several geographies.

Ankit Dudhwewala (CallHippo): Like where we work, Dhruv, is like, so when I said 5000 plus competitors, we have about 4900 of them will work on solving venena problems. What we try to solve is more complex problems. So when I tell you the issues or whatever we are observing, it's more complex. But I will still go ahead and like, I'll try to explain that. So one thing which is happening is that telephone networks and telephone companies worldwide are becoming more and more aware of the spam issues. If you were making, if you're trying to make phone calls and trying to sell through phone calls about five years back, it was much easier. Now, it is continuously becoming more and more difficult. Phone calls get blocked when you call from India to Ireland, Switzerland. And these are the random things which happen. It's difficult to understand. But if you are facing this issue, you will find it very complex. And those are the major challenges which we are also trying to solve currently. And it's coming up like anything. And it's increasing like anything for our customers.

Dhruv Sharma: In fact, you bring up something very interesting, Ankit. I think there'll be tremendous value in sort of going deeper into this. As consumers, sure, all of us get irritated when we get an unsolicited call, but the businesses do have products to sell. So as businesses, what steps can one be taking to ensure that you don't go land in people's spam filters and you're not thought of as a low value product?

Ankit Dudhwewala (CallHippo): No, but there are government regulations like do not disturb. And you can actually mark your phone number as not to be called.

Utsav Somani: Do you think that works? Because I'm signed up for it. And I get probably like 10 calls a day.

Ankit Dudhwewala (CallHippo): At least in India, it doesn't work really well. But globally, it works. If you are in the US and businesses have to follow it, you don't have an option. But you have to follow it.

Utsav Somani: Sorry, you're answering Dhruv's question.

Dhruv Sharma: Yeah. I mean, as a sales organization, how can one stand above the crowd and not be seen as a spammy organization with junk to sell?

Ankit Dudhwewala (CallHippo): So what we help our customer is to understand signals where you can actually call a customer. So we have so much of data now. What we are able to tell our customers is when Utsav will be picking his calls. So we know when Utsav is in good mood and when it is the right time to call Utsav. And when he is not going to mark the phone as spam and get you a telephony complaint. So we can actually go in that depth with data and help our customers.

Utsav Somani: Ankit, your website lists a lot of products in your suite of offerings. So can you describe the ones that you're most excited about in a more forward-looking manner? So going forward in one or two years, where do you think this space has?

Ankit Dudhwewala (CallHippo): So essentially, almost everything combines around communication. We have divided communication into two parts, which is telephony communication and messaging communication. But essentially, it is communication and handled by two different kinds of employees, which is human employees and AI employees. So, but essentially everything is similar and it's one single product which gets weaved together. What I'm really excited about is the AI agents, but not for making outbound calls, but only for inbound calls. So I've also written an article on LinkedIn around this. If you have very small ticket size products and the customer is willing to buy and desperate to buy, or has a very high need, AI agents can do, can sell, but that's a very, that's a niche, which that's, that's the only niche where AI agents can sell. But that in itself is a very big market. So if you are, if you are Pinelabs and trying to sell the credit card machine and a retailer wants to buy it, who has opened a store and is not able to sell, or is not able to increase the sales because he does not have the credit card machine, the cost of product is very low. The cost of product is around 10,000 rupees and the willingness to buy is very high. Here, AI agents can actually, and there's no closing of deal here. The AI agent only has to give information here and the customer is, because the customer is already very, already has a very high willingness to buy. I think that portion of sales is going to be replaced by AI agents. For us, that is a very big market and a big problem which can be solved. And I think next two years, three years, we are going to focus primarily on that problem.

Dhruv Sharma: Ankit, as you said, I mean, for low, low ticket sales, AI agents can receive inbound calls. I think one of the trends we're also observing for high ticket sales and the use of AI is to just give, you know, salespeople personalized coaches so that they can become sales. Do you have a view on that?

Ankit Dudhwewala (CallHippo): Yeah, of course, we have that in Qualipo as well. We have coach where we help salespeople understand what to talk about. We have things like, we will also tell the sales guys, like our sales, our users calls across the world. And when you're talking to someone in the US, you talk 30% about, you do 30% small talk and maybe 70, 80% of your talk is around the product, but you need to have small talks. And the guy who is sitting in India or the guy who is sitting in Australia does not really know which football match happened yesterday and what is the US excited about. So we'll even go in depth and tell our customers that this is something which you can do a small talk about. And all those data comes from AI agents and the coach, Qualipo coach product, which we have. We also tell them what was the last conversation, which had, what is something you can talk about, USP features you should mention at this point in time. If your talk ratio is high, if the sentiment is going bad. So that is another portion of AI agents. Which we work on, but this entire portion is about three, four years, five years old.

Utsav Somani: And maybe as a final closing question, Ankit, there's a lot of narrative that's going around, especially in the world of AI, where founders are being advised by investors and generally the ecosystem at large, where they said that if you're selling and building for the US market or the international market, I think the center of attention and gravity right now is San Francisco for that. So you need to base yourself there to be closer to customers, closer to talent, closer to the market. What's your take on that? Building it out of Ahmedabad, bootstrap. I mean, it's crazy.

Ankit Dudhwewala (CallHippo): I actually like building it out of Ahmedabad. There's a lot of, I mean, we can, we can keep ourselves and I will even avoid Bangalore for that matter of fact. We can actually keep ourselves away from a lot of noise. I mean, and the only reason we have been, I think, able to reach here is bootstrapping is because there was not much of noise around us. All we wanted to do is keep building the business and keep getting revenues. I think San Francisco is good if you are trying to do something which is phenomenally different. And if you're trying to build AI agents and write, like, I mean, build an infrastructure around AI agents. But if you are not really doing that and you are building products for businesses, you can do it out of Bangalore. You can do it out of Ahmedabad. You can do it out of any city, any country in the world is what I think. But if it is really high tech and it really needs those kinds of engineers and those kinds of funding support, that kind of funding support, maybe going to change the world, maybe block ships, like spaceships. If you're building spaceships, you need to be in San Francisco. But if you are not really doing that, I think any country is okay.

Utsav Somani: And you can actually do spaceships now from India as well. I think the space that you just learned is booming in Gujarat and Bangalore and many other pockets. So I think space tech entrepreneurs of India. Yeah, I'm wishing you all the best and see you next week for the offline launch in Ahmedabad. Thank you so much for coming on the show. All right, listeners, we're welcoming our next guest, Ila. Ila from EduFund. Ila, welcome to the show.

Eela Dubey (EduFund): Thanks. What's up?

Utsav Somani: Seems like we're having an Ahmedabad and offline special today. Where are you dialing in from?

Eela Dubey (EduFund): Yes, I am. This week, I'm in Ahmedabad.

Utsav Somani: All right. So let's introduce EduFund for our listeners. It starts off with a personal story for you, because you dropped out of Columbia with a 250K pending invoice for your master's education, I believe. And that's the reason why you wanted to introduce EduFund to the world. Is that the correct story?

Eela Dubey (EduFund): Yes, yes. So I actually did my undergraduate education from NYU. And as a young 18-year-old, you don't really understand the concept of loans, right? And so throughout the entire tenure of my college education, I had these loans compound. I worked two jobs, had a bit of a scholarship, and I made it through. By the time I got ready to do my master's at Columbia, I did have quite a bit amount in loans already. And one semester in, it just did not make financial sense for me to continue. And so I did something which I guess is the antithesis of what a typical Indian student does, is I dropped out. And that left a very lasting memory, I think, on me, which is education, as much as we like to say, is not a privilege. No, it is a privilege, excuse me. It's not a right. And how do we find the solution for this challenge, which is the rising cost of education?

Utsav Somani: And what is EduFund right now? What does it do?

Eela Dubey (EduFund): EduFund helps parents financially plan for their kids' education, with the large and primary focus being helping these parents save and invest. So we help these families reverse engineer a corpus 5, 10, 15 years out, looking what education costs would look like, and then reverse engineer a target plan for them of how much they need to save and invest, and get them to start on that journey. We manage north of $100 million in assets already for a lot of these parents. And then in addition to that, we help parents find additional sources of financing if plan A doesn't work out, right? Because it may not necessarily work out. So we help with education loans. We help, you know, we've put together an entire ecosystem for the families that want to go the extra mile by finding counseling. If the kids want to go abroad, how do they get situated abroad? So it's a full stack platform for education planning and financing.

Utsav Somani: And just from understanding, how much does a Howard undergraduate cost now? And how much will it cost in like, say, 20 years?

Eela Dubey (EduFund): All in, it will cost you about $320,000 from tuition, living for four years. And when we look at inflation, we're looking at about 1.5% in dollar terms year on year, right? And that's not even considering rupee depreciation. So if parents are not factoring in rupee depreciation, they will not only have to deal with, you know, the compounding cost of education and inflation, but also the fact that they're saving in rupee and not investing a dollar, which could actually cost you 70% more on that actual tuition.

Dhruv Sharma: Well, yeah. You know, when a typical undergraduate student, you know, lands in the US to attend a program of their choice, let's say they're going to spend $100 that year on tuition. How do they hack that $100 together? How much of it is scholarship? How much of that is loan? How much of that is savings? Can you paint that picture for us?

Eela Dubey (EduFund): It varies, honestly, depending on family to family, parent to parent. Typically, what we've seen is for the undergraduate cohort of Indian students looking to go abroad, 70 to 80% of that actually comes from family savings. And another 20, 30% of that does come from scholarships or loans. Now, that's actually changing. And it has changed quite considerably in the last 10 years because we're seeing the rise, the prolific rise of our middle class in India, whereby you have more middle class families aspiring to send their kids abroad at the undergraduate level. Historically, that was all, you know, rich or H&I families. But now that, you know, you have this booming middle class that's highly aspirational, that has access to more media, they want to send their kids abroad sooner, not just at the master's level. So now we're seeing a portion of that scholarship contribution, that loan contribution rise as part of the overall portfolio. That being said, parents still need to have saved a hefty amount if they want to send their kids for an undergrad abroad, because it typically still comes from them, their pocket.

Utsav Somani: And are the destinations changing? Because I read a statistic which said that 45% drop in Indian students going to the US, partly because of H1B visa issues. So maybe people are not foreseeing a working environment there, which is favourable for them, or even the cost rising or the geopolitical issues. Are the destinations changing for Indian students going abroad?

Eela Dubey (EduFund): Yeah, there, I would argue in the last year, there was about a 60% contraction. For students aspiring to go to the United States. So yes, there were a lot of changes, particularly due to the geopolitics. Students though are still very aspirational about going to the US. So what we have seen is a lot of students have deferred their plans, right? Because it's not so easy for me to pivot to a UK or a European country overnight, especially in my entire application, my entire aspiration was to go to the US. So we have seen a few of, you know, a few percentage of those kids pivot towards the UK and other European countries. But we also have seen a lot of folks defer with the hope that yeah, things will change in the future. Also, the US has 4,000 universities. There's no other country in the world that can say that, right? And so just the sheer optionality and the diversity of universities and streams is, you can't find anywhere else, right? And so that it's still a top choice for many. But yeah, things are definitely not the same anymore.

Dhruv Sharma: And the students still think it's worth their while to spend, you know, pretty penny going to a non- target school? I know that's a controversial question, but.

Eela Dubey (EduFund): Look, I always say that education is a very personal choice, right? And your education is just as much what you get from it, as much as it is what you give to it and how you make the most out of it. There will be kids that had very big aspirations to go to XYZ University. They didn't get their visa, okay? That is the hard truth of today. A lot of kids are not getting their visa to go to the US. They'll defer or they'll wait. Or they'll have a list that is maybe a potential compromise for them not in their country of choice. There's no one-size-fits-all rule. And if I've learned anything about this market, especially in the last 18 months, is it's just so dynamic.

Utsav Somani: Yeah, just like how everything's evolving in the software world as well. But you've got four parts to your business, planning, investing, lending, and counseling. Counseling being the one where you handhold them in terms of getting to the right choice. How are the businesses performing? Like what's the scale of them right now? How many families or students are you working with in all of them?

Eela Dubey (EduFund): We actually have five. So we just launched a new product within the company called Vittam. And that's a mutual fund distribution arm. I can come to that later. But in terms of where we're actually seeing the most growth, it's definitely one on the savings and the asset management side where we're growing AUM. Like I mentioned, there has been a lot of contraction in the study abroad space. So we have seen the need for loans definitely come down. The aspiration to do counseling, people are thinking about it again. But that being said, we've always wanted to build a very diverse business so that when one thing all kind of comes down, we've hedged. And that's exactly what we've done on the asset side. So we're now at over 250,000 families. Like I mentioned, we manage over 1,200 crores of assets now in that. And despite how markets have been performing, we still see a lot of families come in saying, yes, we want to save because it's for the future. It's not for the now. It's 10, 12, 15 years from now. Out of the five, the AUM asset business, the savings, like I said, is doing the best. And then the other arm of the business that's actually performing pretty well for which we've launched a new product is Vittam. So that's a mutual fund distribution arm. And the reason that we launched something that may not seem very close to what we do is because about three years ago, we realized that the best way to distribute financial asset products and to teach these families about loans and savings is actually through the mutual fund distributor. They are the financial influencer in India. They control where families put their money, how they saved. And so we thought, okay, as part of our GTM, because there's only so much digital you can do, right? There's only so much performance you can do. Why don't we start working with these MFDs or IFAs? And before we knew it, we had 300 partners that we were working with offline and we didn't know how to manage them. So we had to build an entirely new product and platform and ecosystem and brand called Vittam. And that is for our distributors, for the IFAs that sub-distribute all of the products in our stack.

Utsav Somani: So would you classify yourself more like an asset distributor versus an education counseling business?

Eela Dubey (EduFund): I would say we're more fin, for sure. People have always asked us or said, oh, you're an ed business, you're an ed business. Actually, no ed business at all.

Utsav Somani: So that brings me to my next follow-up question, that how does your financial planning differ from say them going directly to a wealth advisor?

Eela Dubey (EduFund): Well, one is it's digital, right? So we've always, we were very clear that we want to use technology. A lot of traditional wealth advisors today, you go to their office. We wanted to reach the masses. So tech was first. The second on the edu fund side, we were always very cognizant that we weren't going to diversify out of any goal outside of education, right? So the way that we advise these families was on the precision of education inflation, on the precision of rupee depreciation, if the kids wanted to go abroad and what that corpus actually looked like and reverse engineer a very specific plan. So the whole bread and butter of what we did was uniformly focused as that, whereas a wealth advisor typically tends to be more agnostic, which is now what VITAM is in some ways.

Dhruv Sharma: And what percentage of those assets get invested or deployed into global assets to generate some sort of return either?

Eela Dubey (EduFund): So we have about 15% of our client base exposed to global investments, either through fund of funds through the AMCs or have direct stock exposure because very early on in our journey, back in 21, we knew we wanted to offer investments in US markets, right? To hedge rupee depreciation. So we tied up with the US broker dealer and now families can come to us and invest abroad. And now with gift city, that's also a lot more accessible and we have various partnerships with gift AMCs as well.

Dhruv Sharma: You know, to me, it seems like you've almost put together this private endowment, David Swenson style, where parents contribute so that students can receive a grant when the time comes for them to go to college. So it's pretty cool. But which brings me to a question, shouldn't the universities have been doing this? Haven't they awakened to the fact yet that, you know, it's just education is expensive. Student debt mounts up very quickly. It's going to change the jobs landscape and there might be, you know, people's relationship with the university system might never be the same again.

Eela Dubey (EduFund): Yeah, I agree with that. But you know, in 2020, when we started the company and we were raising nine out of 10 investors or VCs were like, Ila, the scope of education is going to change dramatically. Everything's going to go online. Life is never going to be the same. And I was like, OK, we'll see. We're in 2026. Your kids still show up to universities. These universities charge a lot of money. Nothing.

Utsav Somani: Only going up like, you know, Steve, I think might be in the lending business where they've survived this long. They'll survive that long.

Dhruv Sharma: Have you guys heard the Thiel statement that insurance is that education is an insurance product, not an investment product? Yeah, I think that he says. I think.

Eela Dubey (EduFund): Because you're capitalizing, and I hate to say this, but you're capitalizing on the fear of parents and the fear of parents is endless when it comes to their kid's future. Right. And it's a horrible thing to say, but it's also the reality, right? A parent will go to all ends to make sure that their kid's future is provided for. I'm sorry, is that me? I apologize. To all ends to make sure that their child's future is provided for. At all costs. So yeah, I know either of you're right. It is. It is an insurance product. Yes, now I don't know. I don't know how things are going to change with AI, but and I'll have to get you the exact name of this, but there is a school program in the US. Someone told me about it and they're using AI tutors and AI modalities to teach these kids and they took kids from like the bottom 10 percentile, like charter schools, the bottom 10 percentile in academia, and they put them in front of AI tutors and AI learning mechanisms for about a year. A year and a half later, 18 months later, they were at the top 90th percentile. How much does this school program cost? I think it was north of $40,000. That's not cheap.

Dhruv Sharma: Yeah, it isn't.

Eela Dubey (EduFund): That's not cheap at all, right? So maybe the way that we learn will change, but I mean, I'm still paying a $20 cloud subscription. I love it, but I have to pay for it, right? I have to pay for access to information and intelligence. I don't think that's going to go anywhere.

Utsav Somani: I think with, I mean, the Khan Academy founder, I think recently said, and he corrected a statement that he's made, I think, three years back where he said that Chad Chibri is going to redefine education, but he said that very recently, it's a motivation issue, I think everyone has access to all of these intelligence tools, but how do you get people motivated to use them in the right way? When do you find it in yourself to actually ask the question that it needs to be asked? But I think what are the other common misconceptions? You've answered the inflation one. You've answered the cost going up and people not accounting for the USD and our depreciation angle as well. But what are the other common misconceptions that Indian parents have when it comes to overseas education?

Eela Dubey (EduFund): But they can wait to start. I think that's the most common misconception all parents have, right? So we don't just deal with the overseas segment, by the way. 70% of the families we deal with are looking to save and start for education in India. The biggest misconception is I can figure it out later. No, you can't. You should figure it out now because when you see that massive price tag, unless you are a C- suite executive at a large company, how are you going to expense for that education, right? I mean, I know a father who sent his two daughters three years apart to a very premier university in the Boston area. For those two daughters, he must have spent at least 700K. Well, in a span of six years, right? That's not I will see later. That's a now plan, right? That's a now plan. That's the most common misconception I see everywhere. And then the other big common misconception is asset classes that I traditionally relied on, real estate, FD, gold will be able to get me the same or similar returns to what I need to match that education inflation. Education inflation is higher than household inflation, guys. 10 to 12%, right? That's a lot of money, right? That's a lot. So to think that traditional asset classes you relied on back then, it's it's a different world. It's a different world. You need to diversify those financial assets elsewhere, for sure. So these these two are the big ones.

Utsav Somani: And as a final closing question, if you were 17 years old, just about to head for your undergrad in Boston, I believe. What is the message that you want to leave to your younger self right now?

Eela Dubey (EduFund): I would say. Just keep at it and it's OK to fail. You're going to fail a lot, but you'll figure it out. And take big risks.

Utsav Somani: Awesome, Mila. Thank you so much for coming on the show. We'll see you for the offline launch in Ahmedabad next week.

Eela Dubey (EduFund): Thanks, guys.

Utsav Somani: All right, listeners, our final segment for today. We've got Akash of Akash Institutes and now Sparkle Ventures. Sparkle is an awesome name. And Akash, thank you so much for joining us on the show.

Aakash Chaudhry (Aakash Institute): Absolute pleasure. Looking forward to the chat.

Utsav Somani: Where are you dialing in from?

Aakash Chaudhry (Aakash Institute): I'm dialing in from Nice.

Utsav Somani: Cannot complain. I did not see you today in the morning, so that's why I was wondering where in the world are you these days. But let's introduce, I mean, everybody knows the Akash story very, very well. But I think today let's focus on the Sparkle side of things. What does Sparkle do in your words?

Aakash Chaudhry (Aakash Institute): Well, actually, Sparkle is sort of solving the problem that I was not able to solve with the Akash ecosystem, which was how do you take personalization and bespoke education to masses? I think in this age of technology and AI, I think we have a very, very interesting opportunity to scale quality education, personalized education, you know, more to say that. With Sparkle, that's exactly what we are trying to do. And we are trying to do that not only for students, you know, across curriculums and across geographies, but we're also able to use the same platform to help enterprises solve their personalized training problems. And I think this is something that we realized in India where multiple languages, there's blue collar jobs, white collar jobs, and training generally gets restricted to a video in a classroom or a trainer in a classroom. People come, listen, have their meals, then goodbye. So I don't know who got how much is always a question in the mind of the promoter, the mind of the trainer. So you're solving both the problems here.

Utsav Somani: And this one statement, which I referred to in my previous segment, also from Khan Academy's founder, Sal, where he said that AI was supposed to be, I mean, a game changer for the world of education. We're still very early in that journey because they're unable to solve the motivation problem. I think everybody has access to these tools, which give you the world's intelligence at a fingertip, but you still need that human element of motivation and also some human touch to this whole thing. And that's why universities will stay around and stick around for a while. Their roles might evolve. And what is your take on human led, but AI assisted education going forward?

Aakash Chaudhry (Aakash Institute): I think that's the way forward. The human element is the motivation, inspiration and the regimentation angle. But the problem that we all had earlier was we were too focused on the content. There was lack of information, lack of knowledge. There was nobody to, you know, reach out in case you need any help. I think that problem is solved by AI, the content problem, the knowledge problem and the assessment problem, which would actually take a lot of time. But now with AI and the human in the loop, you can actually combine the best of both, which is monitoring, inspiration, motivation, regimentation by the humans. And the content, knowledge, assessment and analysis, which is quite deep and multifunctional. So the power of these two has to come forward. And I think that's what we are seeing. Whatever programs that we are delivering, we have a facilitator in the room or facilitator in the system, which is enabling and sort of directing the user or the student or an employee as to how to maximize from this content and also ensuring that, you know, their discipline, they are coming back to it and they're actually ending up finishing the programs. So that's the power that we feel. That's the, I would say, the configuration or equation we feel is going to go forward in the age of AI.

Utsav Somani: And is it one to one at Sparkle or is it one to many or one to smaller batch?

Aakash Chaudhry (Aakash Institute): We started with one on one, which was actually a teacher led. But when we finished about 50,000 hours of this entire interaction with the teacher and the student on multiple subjects from multiple geographies, we are now teaching in about 20 countries from India. So that's the knowledge that we sort of, you know, sort of inserted and injected into our systems. And that's where we, you know, got a deeper understanding of what does it take for an AI to give instructions on a one on one basis and how do they handle the exception handling and the questions that are coming from the other side. So what kind of language, what kind of empathy, what kind of depth and how to assess, you know, the quality of the question. So all of that came from our actual live tutoring that we recorded over a year. And then we started with one on one in a live format. And then we took that to one on one on an AI format. And we also made it one to many with a facilitator where there's a facilitator that is ensuring that there is a discipline in a classroom while AI does its job on the white screen board.

Dhruv Sharma: Akash, personalization is relatively easy to understand in a tutoring format when you're receiving one is to one instruction, which was a big part of the Akash story also in the early years. But what does personalize mean inside of a classroom? And so what I mean by that is, you know, in the manner in which we were instructed in school, time was the constant. We had to get through the syllabus in three months or six months. And comprehension was variable at the end of the semester or the term. Some students understood the matter better than the others. With personalization, I think there's an opportunity to flip time to be the variable. It doesn't matter if it takes less time or more time, but everyone exits the classroom with 100% understanding, no gaps in their knowledges. Is that the direction in which we can really push education?

Aakash Chaudhry (Aakash Institute): In fact, you know, that's the holy grail, because every child learns in a very different way. You know, some people are auditory in their learning system. Some people are visual. Some people are kinesthetic. I mean, they want to write and learn. So everybody has their own sort of, you know, method of learning. And this is what we realize, that you can't just, you know, put them in one classroom and expect all of them to understand at the same level. And what AI is able to do that is actually break that barrier where they say there is no further time limit. You don't go ahead till you get the concept right. And I think that exactly is what, you know, we are trying to solve in Sparkle, where you can continue to delve into a topic for minutes, hours, weeks, till you get it. And I think that's what we're able to turn around, not only for, you know, students, but also for employees, that, you know, till you don't get all the terms right in the language that is comfortable with you, you don't just go ahead.

Utsav Somani: And Akash, I think what was interesting was that you were, I mean, the students that you were admitting were almost adults back in the day. But now you're going after a much younger audience. So I believe the question for you also involves and any product that you build has to be sort of in consultation with parents as well, because they will have to be involved in delivering this education to them also, or at least the platform that you're building for them. How has that role evolved and what are conversations like with parents these days when you tell them about this offering?

Aakash Chaudhry (Aakash Institute): Oh, you know, it's interesting you ask this because, you know, in Akash also, the age group that we were dealing were, you know, right from 13 years to 18 years. That was the sort of the age bracket. But here we're doing grade 6 to grade 12 and beyond in terms of test prep that we handled. But I think one of the key reasons for us to be successful in Akash was we never, you know, kept the parents out of the loop. So we always felt that it was a joint venture between the institution, the teacher, the student, and the parents. So these are the four key pillars on which the, you know, the success stories were built. And that's exactly what we replicated in the Spark ecosystem. Even if we are doing an AI, we are doing a one-on-one. There is a massive amount of interaction that we are doing with the parent because, you know, they are giving us a real feedback that what you did with my child, what am I observing at home? Because that, you know, goes as a feedback loop into our system. That, you know, is the learning journey that we curated for this child, you know, throwing outcomes. And in education, you know, you may bring the best of technology or the best of the teacher, but at the end of the day, if you don't get the outcomes, you know, there is no use for it. It's like, you know, I mean, you may have the best of the best, you know, cars or planes if it doesn't take you to the destination, it's of no use. So I think that's what has been our driving force, that how do we bring the outcome and to bring the outcome, understanding the student and the parent, both is very, very important.

Dhruv Sharma: And Akash, you know, on the show, we've been talking to people who are bringing AI into the clinic. So making a clinician's life easier so that they can do what they love doing, which is curing patients. And teachers love teaching, you know, but part of teaching is also doing all of this other stuff, like grading papers and a whole bunch of admin work. So when you travel and talk to your, you know, your peers, other education leaders, how is AI making the life of teachers better so that they can focus on the core mission of imparting a world class education?

Aakash Chaudhry (Aakash Institute): You know, this is very, very important because one of the things that was keeping the educators away from delivering high quality education was all these administrative and mundane tasks, which AI can handle very, very well. And outside of that, you know, I'm sure you can relate to the phenomena, you know, in your own situation where there is a certain kind of bias that comes in. You know, if I've had a way in which I've made a kid successful, that I would really stick to it for my life. And this is the way to make children successful. You know, so there are certain biases on that. If I'm assessing an assignment or an exam or a test, if this is the way, you know, I assess, these are the qualities that I look into and somebody tries to do something different, I'm not going to accept it. So I think AI solves for all the kinds of biases, all the kinds of repetitive work or a mundane work that would actually bore people out or would not make them deliver at their best levels of performance. So I think that has been a massive advantage that we are seeing from AI as far as administrative, repetitive, boring work is concerned. It is doing it with same rigor, the same assessment quality that you feed into it. Without any lapses. I think that's something great. I don't know if you came across this recent CBSE exam mark sheet checking where even the right answers were marked wrong and given lesser marks because, you know, the teacher was doing it manually and you get fatigued, you know, assessing thousands of papers. So how do you keep up? But AI can do that beautifully.

Utsav Somani: And there was one story that I read recently which scared me. Like, I think somebody hacked the CBSE servers and they were about to edit like names of teachers and stuff and they had to report it to the board just as a white hacker. But it's scary. I mean, Anthropic Mythos is released on our Indian legacy systems. We're in for a thing. But how much of Sparkle is enabled by AI? Like if Sparkle is like, say, 100 units, how much of that is AI right now? How much of that do you foresee in the future will be AI? And is Sparkle possible only because of AI right now?

Aakash Chaudhry (Aakash Institute): So the current format that we are offering in the market in terms of the tutoring is one-on-one teaching. So only that part is human. And outside of that, you know, the entire backend in terms of, because every video, every lecture that gets delivered is recorded. And, you know, the videos are segregated, audio is segregated, text is segregated, and they're all fed into the algorithm. And that's what it decides for the next session that the child should take. So if you were to give it a number, I think on a one-on-one, there is, I would say we would be about 15 to 20% AI. But when we move on to the AI tutoring and AI training, that is 60% AI. 40% is, you know, content generation, content review, back and forth logistics, uploading, ensuring the feedbacks and everything. But that we feel going forward will also inch towards a very high degree of automation. But on the education side, we still are at 20% and we hope even if we achieve 50% it'll be a massive, massive advantage.

Utsav Somani: So maybe a thought experiment here. We already have AI Companions. We had a founder on the last show who was building AI Companions, mostly through chat and text right now. But given how the video models are evolving and how they can read emotions now and voice and everything is evolving to a point where you might actually be talking to somebody on the opposite end who's literally better at reading your emotions and answering questions in a more neutral way than even a human. Do you see that happening where in three years, in 2029, I'm sitting, taking classes from somebody who looks and feels and talks like a human and understands me like a human but is actually not a human teacher?

Aakash Chaudhry (Aakash Institute): Three years is a long way ahead. I think you will see that in probably a year. We've also internally experimented a lot of models and in fact on the enterprise side, we've got role modeling where people can do a role play that I'm a teacher, okay, you ask me a question and I'm going to tell you where you're going wrong and how I'm going to help you navigate. We just have to put a face to it. Right now, the technology isn't that accurate, but I think given a 12-month timeline, we would be actually looking into some of these solutions, not in an experimental form, but actually in a go-live production form. That's how things are changing, that's how fast things are changing.

Dhruv Sharma: I just want to try and get an insight from the time you spend on the developmental and the philanthropic side, right? Which is, there is a real risk that AI might even further widen the what they call the digital divide. So, for the vulnerable population, what is it that we're already doing? What is it that we can be doing to, again, in a sense, as Nandan Elkani says, diffuse AI and get in the right hands where it matters?

Aakash Chaudhry (Aakash Institute): I think that's a very difficult and ethical question because unfortunately, a lot of AI efforts are going into building efficiencies and productivity, but the direct outcome of that is lesser need for people. And if there is lesser need for people, that directly takes away the mundane jobs at the bottom level. And that's exactly where this larger part of the population is getting readied. And I think that's the scary part. I think to Nandan's point, if you talk about AI diffusion, I think it's an idealistic expectation, but if you have fire in your hand, you can burn a house or cook food. I mean, it depends who's handling it. And I think that's one of the biggest fears that we have, if not dealt carefully, AI could actually further escalate the digital divide that we see today in a much, much more concerning way that I would say.

Utsav Somani: And Akash, you're involved with many educational institutes as well. You built Akash Institutes and Ashoka University, Plaksha University, ISB and now Sparkle. How is the definition, how are you thinking about the definition of good education then versus now and maybe three years down the line?

Aakash Chaudhry (Aakash Institute): I think the biggest, the challenge that we've faced in the past we could not scale high quality ecosystems. When I say most people tend to believe that a good teacher is the end of a good quality education ecosystem, but that's the beginning. I think a good quality ecosystem includes a good teacher, good resources, good environment, an amazing peer group, and a culture of experimentation and research. So there is a lot of things that go behind in building that kind of ecosystem. And unfortunately, in the earlier era, those were all limited to very few places. But what we are trying to do with AI is that we are learning from it that how do you create those centers of excellence? Is there a model and a framework? Can AI plug into some of those repetitive and automated processes there and scale that? And I think that's the future that we're looking at, with AI actually playing a very, very important but an integrated role with humans to bring that excellence, which was sort of limited or restricted to some pockets here and there, but now is available to the larger masses.

Utsav Somani: And how is your view on competitive exams and competitive entrance exams in India change specifically?

Aakash Chaudhry (Aakash Institute): Oh, I don't know. You know, I really think they should change it. I think they want people to think and be open to ideas and be curious throughout the year or throughout their education journey. But at the end, they want them to sit in front of a paper and answer a question for an exact answer. I think that's where the problem is. None of the business owner, you know, sets up and builds their business in a single way. Everybody has their own way, their own DNA. And similarly, the education also needs to be very, very individualized to the student or to the learner. Right now, we are just testing them on their ability to perform on a particular day for a particular time on a restricted holiday or restricted timeline to check whether they can deliver the same answer as 2 million people. I think that is a flawed model. I think every answer is acceptable. The problem is that who's going to assess in the context of the question asked. And I think that is the problem that led to this mass test prep kind of ecosystem because there weren't so many teachers or methods to assess every child the way they are. It's just simplified into a single test. I think that has done its job. I'm not saying it's bad. It's done its job. But as we move forward, I think we'll have to really transform or reform the way we are assessing our children or grading our children.

Dhruv Sharma: Maybe AI has a role to play there in getting to a place where it can objectively evaluate at scale. Akash, I want to ask you a question as a parent, as a lifelong educator, which is, you know, right now at this point in time, we're surrounded by AI all the time. And so for you to come in touch with your non-AI side, you have to literally keep the tools away. Do you think schools should have like AI periods and non-AI periods or something they do so that students can still learn what it's like, still experience what it's like to ride the cycle without the training wheels, so to speak?

Aakash Chaudhry (Aakash Institute): No, absolutely. Because, you know, right now when students are coming to schools or any education institute, the institution sort of assumes, you know, when they are in a class, they will interact and they will get social, they will learn from peers, they will learn to respect. So those are the, you know, softer culture-based teaching methods that they are relying on. But I think going forward, a lot of knowledge can actually come from AI. And the schools will have to put a concerted effort. OK, you know, let's have a social and a cultural integration session. Let's have a respect and adulation session, you know, with people. So I think, so they will have to really put thoughts now and structure to things outside of the content and knowledge, because that's been solved. Earlier, their massive amount of effort was, you know, what lecture I have to take, for how long, what question will I ask, et cetera, et cetera, et cetera. I mean, the entire bandwidth was going there. But now it doesn't have to. It has to go now on the entire social part of it and the education part of it, the freedom part of it, rather than just the knowledge and the content.

Utsav Somani: And I'm going to lean on this a lot more in the very near future, but I'm a new parent. And if you were to redo education for your own kids, what would you do differently? Are there any exciting new models in US, in SF? Like, I think homeschooling is a big one. Outschool, I think, is a company that a lot of Silicon Valley billionaires are funded and are using. But anything that you think should be experimented more in early years of education for kids?

Aakash Chaudhry (Aakash Institute): I think in India, formalized sports integration into education has always lacked. I mean, people have played it as part of their entertainment, as part of their, you know, socializing, not as part of an integral part of education. Because sports teaches you, you know, physical fitness. It teaches you team building, teaches you how to lead the team. It teaches you how to get up from a failure. I think sports have been completely missed out and left to the children, OK, you know, go figure. If there is a ground, go figure out what you want to do with the free time. Whereas, you know, children I've seen who've been really integrating sports into their day-to-day lives, I think they have, the kind of resilience that they have built is very, very different than what we see kids who are only stuck around with the books. And the kind of world that we're moving in, as they call it the Barney world, which is brittle and sheer, nonlinear and incomprehensible, you would need a lot of resilience there. I think sports is one and allowing, you know, children to, you know, bring in other forms of education, be it music, be it art, be it literature. I think somewhere in the race to become successful in this tech world, people have really ploured their children into technology education and actually sort of disregarded the role of art and culture and literature, which is now coming back. I mean, if you can't prompt properly and articulate properly, you can't get the right answer, right? You're coming back to the communications now. So I think these are the things that if I get an opportunity again, I would really love to integrate these two things deeply into, you know, the educational system.

Utsav Somani: Thank you for sharing that. Dhruv, any final closing question before we let Akash go?

Dhruv Sharma: Akash, what do you spend your time thinking about these days outside of all of this stuff that we've covered?

Aakash Chaudhry (Aakash Institute): Oof, so many things. I think a lot of things, one central theme that I kept and I keep thinking a lot about is, you know, one of the regrets that I have in building the Akash story was, I think I was, I did not think big enough. You know, I marred myself into, I had so many chains around my mind. This is doable. This is not doable. Let's not go there. Nobody has done that. So I think that, I think I should have. And now what I'm thinking now is how do we induce this open-minded culture in the entrepreneurs, in the students, in the parents? Why are they getting so restricted and so sort of boxed into what they have learned or they have seen? They want their children to be safe, but I really want them to also, you know, you are the safety net. If they fall, you're there. So let them experiment. Let them, you know, explore the world today. And then the kids will surprise. And this is something that I keep thinking that thinking boxed and thinking small has been a problem for me. And I realize that now we are thinking in a very different format.

Utsav Somani: Awesome. Akash, thank you so much for coming on our show. But before I let you go, there's a little birdie that told me that you're wearing the Deepinder World temple device these days.

Aakash Chaudhry (Aakash Institute): Oh, yes. I was experimenting with it. Yeah, it's a new, it's a new device. I actually enjoy that form.

Utsav Somani: Yeah, what can you tell about the product in itself? Like any, I mean, any things that we can learn from you?

Aakash Chaudhry (Aakash Institute): We've just started. So we're just calibrating, you know, how it sort of reads my body stats and everything. Still in the early days, but looks promising. Maybe, you know, Deepinder says that, you know, they will experiment with the form and the placement of the device as they go forward. And looking forward to, you know, what this device brings up.

Utsav Somani: All right. Thank you so much for coming on our show. Enjoy the rest of the day.

Aakash Chaudhry (Aakash Institute): Thank you guys. Enjoyed.

Utsav Somani: All right, listeners, that's it from us. We'll see you on Friday at four o'clock. Thank you so much for tuning in for our live stream. Episode number 95. Bye-bye.