Edtech Insiders
Edtech Insiders
90 Million Users Strong: Kevin Leem Discusses the Success of QANDA
Kevin Leem is the CFO of Mathpresso, operator of the globally recognized AI-powered learning platform, QANDA, which serves over 90 million users. QANDA provides AI tutors that offer personalized support to students anytime, anywhere, enhancing their learning experiences. The platform also equips teachers with AI teaching assistants to improve teaching efficacy and student tracking. Driven by the belief that personalized, high-quality 1:1 tutoring is the most effective form of education, Mathpresso is dedicated to making such experiences universally accessible through technology.
Kevin brings strategic and operational expertise from his roles at Morgan Stanley Private Equity and Credit Suisse, backed by a strong foundation in Finance and Accounting from The Wharton School at the University of Pennsylvania. He is deeply passionate about using technology to overcome educational inequalities, aiming to ensure that all students have access to top-tier educational resources, irrespective of their financial or geographical circumstances.
Recommended Resources:
📰 GSV: AI & Education by Claire Zau
This season of Edtech Insiders is once again brought to you by Tuck Advisors, the M&A firm for EdTech companies. Run by serial entrepreneurs with over 25 years of experience founding, investing in, and selling companies, Tuck believes you deserve M&A advisors who work as hard as you do.
Kevin Leem is the CFO of Mathpresso operator of the globally recognized AI powered learning platform QANDA, which serves over 90 million users. Cuando provides AI tutors that offer personalized support to students anytime anywhere, enhancing their learning experiences. The platform also equips teachers with AI teaching assistants to improve teaching efficacy and student tracking. Driven by the belief that personalized high quality one on one tutoring is the most effective form of education math Presto is dedicated to making such experiences universally accessible through technology. Kevin himself brings strategic and operational expertise from his previous roles at Morgan Stanley private equity and Credit Suisse. Backed by a foundation in finance and accounting from the Wharton School at the University of Pennsylvania. He's deeply passionate about using technology to overcome educational inequalities, aiming to ensure that all students have access to top tier educational resources, irrespective of their financial or geographical circumstances. Kevin Leem Welcome to EdTech insiders.
Kevin Leem:Hey, thanks, Alex EdTech Insiders for having me and featuring QANDA in our new product graphi on the show, and it's absolute pleasure for us. Thank you very much.
Alexander Sarlin:Yeah, we hear a lot about math, espresso and QANDA, from the US edtech market. But I would say a lot of people in the US don't know a ton about it. Even though it's absolutely massive, you have 90 million users, you're the number one sort of edtech tool in Asia. And you're out of Seoul, South Korea, but like because we haven't had much exposure in the US. I don't think people know a lot about it. So let's start by just telling us about QANDA. And, you know, how did it start and give us the kind of story?
Kevin Leem:Sure, so hopefully, I'll be able to provide with you some the inception of QANDA, and maybe you can build more on like, how it operates and all that. So I mean, before we start, I would like to give you a bit of context on education environment in Korea first, because it all makes sense from there. So, in Korea, getting into prestigious university is considered like the most crucial key to climbing the social ladder, as its competition is very, very fierce. And because the public education alone cannot cater to each student's need a level. As a result, most students rely heavily on private education after school such as cram schools or private tutors to receive extra study help. But however, the quality and accessibility of private education vary significantly between so and Nonso areas, like due to economy factors, and all that so, and the co founders upon the are all non sole residents, and they experienced all these problems firsthand. And they really did struggle to find good tutors or the hometown while they were studying for their college entrance exams. And one of the co founders Ray after he went to college started working as a private tutor to pay for his college tuition. And he soon realized the stark difference in educational resources between the previous Seoul areas and other regions. And just to give you like, a quick example, like a handful of the students from certain district in Seoul had multiple tutors for one subject. And sometimes they're even getting tutored, to better understand lessons from another tutor or cram school. And but I mean, what's it's extremely difficult for students in other rural regions in Korea to find one single good tutor. So Ray ended up sharing this experience with his high school friend Jake, which happens to be another co founder of QANDA. And they came up with this idea of a mobile q&a platform and hence, that's the word the name QANDA came from, which is to connect students nationwide with high quality tutors. And they did this to provide accessible and cost effective option for C units to study more effectively, and also providing extra source of income for tutors. Yeah. So that's the genesis of Kwanzaa. And hopefully it helps to understand like how we started to build this platform.
Alexander Sarlin:No, it's really, really helpful. I didn't know that quando actually came out of that inequity. That is obviously sort of structural in Korea and South Korea. I remember reading Amanda, Ripley's terrific book about education called the smartest kids in the world. She talks about education in different countries. And one thing that always stood out to me, she talks about Korean education and how a lot of students will be sleeping in class. And the sleepiest students are actually the ones who are working hardest, because they have all these tutors. They're doing all this extra private education. And when they come to school, that's sort of like gravy. So they can sleep through their regular math classes and STEM classes, because they've been tutored at home. So it sounds like that's a huge inequity problem. And it's, as you say, it's urban and rural. Right, right. So that's the inception. And it's grown enormously. Tell us a little bit about how cuando has sort of evolved from that inception. It's huge now, but I'm sure it took a little while for people to fully understand it and get their head around it. Yeah. So
Kevin Leem:there has been a big moment for us to become where we are right now. So we started as a q&a platform, which is simply matching tutors, with students who are in need for Q and A's. But what we figured that the app is not so sustainable, because it's just started off as mere pure matching platform. So the students have repeatedly asked the same questions, and teachers have to answer them all over again. And we realize it's not sustainable thing. And it's just cost students and us too much. And it really took a longer time for the students to get the answers right away, when there's like answers ready out there that has already been answered. So in early 2016, so we decided to pivot our products, and we converted conda into an image based search platform utilizing AI technology. So now the students can actually search rather than to ask questions by taking a photo of a question. And our OCR technology, which is based on deep learning, will find the similar or exact match from our database. And this, we provide it to the students in a way that is most beneficial for their learning process, whether it be like step by step solution, or more of a video type of answers. And we figured this is a lot more efficient, because students no longer have to wait like 10 to 20 minutes to get responses from human tutors. Now, because they only took a few seconds to get the research results right away. And we this is the moment where our app started become viral, and our traffic took off. And we went from receiving maybe 1000 questions per day. So more than 4 million questions to help and sometimes 8 million during exam period time.
Alexander Sarlin:Wow, that is a very high rate. And he was amazing. It really is. And you know, I think some listeners may recognize that this optical character recognition approach to tutoring, you know, cuando was there very, very, very early along with the US company Photomath. Yes, but at this point, because AI has evolved so much this is becoming, you know, more recently, other companies have started to move into this. But this was a hard thing to do. He took very sophisticated technology to do that level of optical character recognition and immediate answering in 2016. How did your team sort of, you know, I'm not asking for, you know, tech secrets, but it feels like, you know, it's a change, but it's a change that needs some technical prowess. How did your team manage it at the time?
Kevin Leem:Yeah, so it was during 2016. And our co founders watched a AlphaGo match with one of the, I guess, the best Go player in the world called Lisa doll. And at that time, everyone thought Lisa doll will win all the matches, but AlphaGo ended up winning four out of five matches. And that was quite sensational, especially in Korea, because this is also Korean guy. And that's when everyone started to put their eyes and focus on deep learning. And start to feel like this is the way that we can pivot our products by utilizing our database and converting into a search image based search platform. So because we have our database, and we will use our deep learning technology so that our model will be trained to see all the answers and question datasets so whenever students ask any question, even if it's a similar question, we'll do a similarity check and try to find anything that will be helpful for students we will provide multiple sets so that even though they it's not the exact match, they will at least get how the logical process of solving each questions. So it starts improve because students started to using more of the app. And as we have more database will pre flywheel and increase the search success rate and which in turns making better of our OCR and the AI technology that's built behind the backbone
Alexander Sarlin:That makes a lot of sense. So early on, you had a training set of problems to solve, but you were flexible, it sounds like a little bit, you know, the idea that somebody might take a picture, and it might not be the perfect match, it might not in 2016, you know, to be able to give the exact walkthrough of that exact question, it would give questions with the same concept, or questions that were similar. And then, of course, when you go from, you know, 1000s, to millions very quickly, your dataset gets really big and your training set gets really big. It's really interesting to hear you talk about the AlphaGo moment, because that was, you know, I think, a shot heard around the world in terms of the AI revolution that the AlphaGo was related to, you know, what became Google DeepMind. And it was this sort of moment when Yeah, I mean, go is in a much more complex game than chess, actually. So this was sort of an exponential jump from what I had done in the past. And it was very prescient of your founders to sort of see it as a cross the Rubicon moment when, okay, AI is here, and it can do amazing things. And we should really do this now, which is great. Before we go on, I mean, we're talking about the technology. But let's go back for a second and just talk about the impact. You know, we're talking about 90 million users, that's a lot of users. And I'm sure it's been growing. What are some of the impacts that you've seen? I mean, has there been a sort of beginning of a reduction in inequality between Seoul and the non Seoul areas? Have you had students write to you and say, I owe my life to you, because I got into the school, I wanted to get into, like, what is sort of the impact look like? Yeah,
Kevin Leem:I mean, definitely the social impact will be have observed, and that's really something that gets our employees cuando keep going. And that's something that is really, I guess, helpful for us to continue to work harder to make this product better. I mean, we receive a lot of feedback across a global student users that conduct did help reduce the financial and to ecological constraints in accessing this private education. So I mean, we have a few examples. I mean, in Vietnam, so the students have to commute 20 miles to attend a private academy there. But he started using Honda in Vietnam, and he no longer has to commute anymore, because contact will be able to solve and provide the personalized education experience for them. So he sent the thank you letter, and saying that he no longer had to commute that for, and I guess a similar feedback that we see from Thailand, he explained that in Thailand, most education services were catered to wealthy students. So the the education inequity is slobbers. From what we have observed, and since conda, is three, it quickly gained popularity across all students. And we've seen a lot of students start purchasing our subscription solution on Ponda, which really made conduct the first educational app in Thailand to achieve mass adoption. And our contact was not only welcomed by students, but we have received a lot of positive feedback from public school teachers, because codnor really helped lessen the learning gaps between students. So my last example is about Korea, where as you said, as I already mentioned, the student to teacher ratio is very high. And students, I don't know why, but tend to be too shy to ask question in the classroom publicly. But because we QANDA, they can ask questions, anytime. So the learning gap between students starts to fall, and it became a lot more easier for teachers to provide the lecture and mentor the student because it keeps the students at a similar level. So we receive a lot of number of orders from teachers for group purchases over a premium version for their own students. So there's a numerous, I guess, feedback like this, and this is something that yeah, really helped our team to, I guess, be thankful about what what we are building, I'm
Alexander Sarlin:sure, yeah. And you know, it didn't even occur to me, in your original Inception story, you're talking about soul and the surrounding areas. But of course, Korea itself is surrounded by countries that are much poorer than it that have less sophisticated educational ecosystems like Thailand and Vietnam. And I'm sure it's been transformative for the entire region, which is really exciting. So that's really interesting to hear. So, you know, one thing that we should make sure to note here is that, you know, cuando has been a big deal in Asia for almost 10 years now. But you just launched your first us specific product, and it's called Krama phi. And it's being piloted right now in California, tell us about what Chroma phi is and what you sort of expect to see as you really break into the US education market.
Kevin Leem:Thanks for asking that question. So Honda, which has been a lot of traffic, and I guess the positive feedback from users in Asia. So clarify, I guess, is a new product of ours. And it's Gen AI native study tool, which is catering to the higher education. And we started out as a product that will help the US college students at first, and this is because as I said in the beginning, it's a genuine native app. And as everyone knows, the foundation model that goes behind it, it works best in English because it's trained mostly in English. First language and other European languages. So that's why we started to introduce in the US market first and especially after getting all the traction from Asia made obvious sense for us to utilize what we have already built turns database and traffic and brand equity to bring it to the biggest market in the world, which is the US. So it quantify AI, what it does is, it's a study tool for exam and coursework prep. So students can upload your study materials that they receive from each classes. And we provide personalized study guide and prep packages. So let's say you're taking finance 100 from University have something then you can upload lecture notes, or videos or whatever study materials that will help you study. And we will provide a, it's basically a tool that summarizes the provide study guide, and we will match the key terms with the original materials that they uploaded. And it will be able to generate like flashcards and mock exams, and all the other helpful tools for them to be able to study for their exam. So we started out having few of our staffs in LA, because that's where we thought we'd have a lot more campuses around that region that wouldn't be able to do a PMF test and having the user interview. But what has been really exciting for us is that as of June, the traffic composition has shown that US accounts for less than 50%. Now, this has already become something that is used by the global users. And it was really exciting for us because we've seen users from like Germany, Switzerland, and Hong Kong, Australia coming in to use our product. And thanks to the foundation model of the Gen AI feature, I can understand all the basic all the languages and because we have to own like materials that we have already built upon the it has actually been really great. And yeah, so we're not targeting us solely at the moment. But we will continue to make it so that we will be helping the US college exams first and approach to a niche market before we start expanding more actively globally. That's
Alexander Sarlin:really interesting to hear. And so, I mean, it strikes me when you talk about how in, you know, in 2015, there was this shift in condors original concept of matching for tutors, to actually let's use the economies of scale here, if you have lots of students asking the same question, why don't we just, you know, train something to answer it really well. And then they don't have to wait. And then of course, in this generative AI moment, that sort of becomes supercharged, because as you say, a student can ask a question in any language get an answer in any language. And the question doesn't have to be one that has been asked before they could ask almost anything. And generative AI is incredibly sophisticated at putting together these massive datasets and synthesizing them into answer. So it feels like a real sort of second generation version of some of the AI that has been fueling condos, you know, massive success in Asia. I mean, I'm excited to see it coming. And I think, you know, one thing that is also particularly interesting about QANDA, and you know, the company behind it is correct me if I'm wrong, the company behind it is called math spresso. Right, right, in Asia, is that right? So
Kevin Leem:met Presto is the name of the company. And conda is a product name and a famous kind of, yeah,
Alexander Sarlin:the reason I bring it up is because, you know, Matt spresso has had some really high profile investors in its, you know, eight year rise. And one of them is notably Google. And we often see, you know, Google back startup mass spresso, as one of the main ways it's sort of covered in the US media. And that makes a lot of sense in a lot of ways, right? Because as you said, you were sort of founded on this go AI revolution. I'm sure Google has had their eye on you for a long time. Google also purchased Photomath in the US a few years ago. So that was a pretty notable acquisition of something that had some similarities to what cuando is doing in Asia. Tell us a little bit about what is it been like to have Google as a prime investor in math spresso.
Kevin Leem:So since day, one of their investment, I think it was back in 2021. They have been one of our biggest supporters of conda through various means. So we were collaborating in their efforts on Google vision. And we been using their GCP cloud services. And even nowadays, because like Gemini is one of the key leaders in this foundation model like products in the world together open AI and other other products. And we are having that private access to Gemini, and we are utilizing that in our quantify so they were helpful in terms of like product and the infrastructure. And also they were a, I guess, a key advisor who would really like reach out to when we are discussing about like the product development and even expansion into the US the other markets and yeah, they were always out there and they You're quite supportive of what we have been building. Yeah,
Alexander Sarlin:I mean, Google's obviously one of the most sophisticated technology companies in the world, but quite a while, and really has a pretty big footprint in education, technology in a variety of ways, as well. So it says a lot that they have sort of put so much faith in Mattpresso, and QANDA. And I think that's, it's really meaningful. I think, you know, some of the listeners here who know, the ad tech industry really well will spot that, you know, when Google sort of believes in you, it means there's something going on, that's something you're doing really, really right. And the connection between, as you say, Google Cloud, Google Gemini sort of creates an entire ecosystem that conduct can really use as a foundation for its growth, especially, for sure. So you mentioned that you get a lot of group orders from teachers who are looking to sort of even the equity gap within their classrooms and make sure that everybody has access to really high quality materials. And that's really interesting that the idea of teachers, you know, actively seeking AI based educational technology tools, because we're in a moment in the US where we're trying to, you know, get our heads around, especially this new gender of AI technology. And we keep getting these polls coming out. Some say that, you know, the majority of teachers have tried AI other say, only, you know, 20% or less have tried it, and some say they're really worried about its impact in the classroom. And there's just a lot of confusion in the space. I'd love to hear you. I mean, do you feel that the teachers who have come to QANDA for solutions, you know, have you ever gotten any concerns about the AI or or about the sort of integrity? Or do they recognize immediately that this is so beneficial for students learning? Yeah,
Kevin Leem:I mean, those kinds of challenges are definitely something that we have had since inception. I think it's it's an thing that challenges that a lot of the EdTech tools have been getting, not only in Korea, but other parts of the world. But yeah, I mean, it's, I think it's two sided coin. So it's up to how well you use these tools, whether it be just AI or general AI types of tools. So if the students decide to manipulate conduct or any similar services for simply like cheating, per se, then it would can be something that would be I guess, frowned upon by teachers because of this integrity issues. But that is one of the few very few like feedbacks that we get, but most of the feedback that we get, is about how it really helped lessen the gaps in education and equity. And I think it has been proven from how we have been performing in the Asia parts of the world where the infrastructure is not there. But the fervor for education is there, like especially in Vietnam, or Korea, where the Asian students tend to be like shyer than the Western countries. And it's really hard for them to find any, like quality tutors, because there's only limited numbers out there. And I think this is something that can be applied to a lot of products. But yeah, I think it's really up to the students, and how they use it. And what we have been a feedback from the teachers, or it's a beneficial tool, but as similar to any other tools, we should still be cautious on how, and we should be mentoring and guiding students how to use them. And we are actively, I guess, talking with a lot of teachers, both in public and private schools in Korea and Japan to help conda app to be a more of a beneficial tool. So we are trying our end to provide more of a like step by step solution type of answers rather than providing answers all at once, so that they can really handhold them throughout the logical process of learning. And also at the same time, we will be working together with like the private institution of gathering all this q&a That been asking by the students of a specific classroom, and we will be transferring that transaction to the teachers so that the teachers will be able to analyze and know what the weaknesses of each students and which are the gap. And so I guess that's like the beneficial part of how you can use less our solution.
Alexander Sarlin:That's actually a perfect segue to what I wanted to ask you next, which is, you know, we see a pattern sometimes in edtech, where, you know, something is originally created as a consumer app or as a, you know, direct individual application. And then there's a realization that, oh, this is being used in schools. This is being used in educational institutions of various kinds. We should introduce some features that will support teachers support school administrators support people in sort of understanding how the students are working and find, you know, patterns and gaps in knowledge and things like that. We Khan Academy is a notable one of these in the US. It sounds like you have done that already. And conda has sort of some teaching assistant type features to help a classroom actually use it, most of it At Tableau, can you tell us a little bit more about those.
Kevin Leem:So we believe there is a crucial element in education, that technology would never be able to replace human tutors, teachers, I think it's interacting and motivating students. And so to address this, and also to empower these teachers to do what they can do best, we introduced a product called Ponza. Al. It's a, a tablet based adaptive learning solution for students, and also a LMS for teachers. So from a teacher's point of view, what it does is that it handles all the tedious tasks such as greeting, generating homework, or generating the reports for them to understand what the student's strengths and weaknesses are. So it allows teachers to dedicate more time to interact with students and focus more on providing the best lecture that they can give. And from the students point of view, this contact al solution provides adaptive learning experience on tablet. So the study materials and problem sets really does like alter based on the learning progress of each student. And because we believe we know and such iterative exercise is an essential part of studying. And this is all done purely via through a tablet solution. And it's generated automatically. And this which was before all done manually by a limited number of teachers. So this is something that we have been introducing. And we have been working together, there's a government led project in Korea called digital textbook, so is trying to transform all the physical textbook onto tablet. So that's something the Korean government is trying to do 2025. So we're participating in that. And in the private education segment, we are working with the private institution in Korea. And we have actually set up our own hybrid Academy called panda Institute, which we leverage this Kwanzaa al tablets solution. And we have actually, we only opened up like late last year, and we've seen major progresses and students like rates, and have received a lot of positive feedback that students actually enjoy this process because they get instant feedback. And these, like learning materials change by the learning progress, without having to wait for their teachers to provide feedback. So teachers point of view, we believe this powerful TA and from student's point of view, I think it will be it's like a powerful AI tutor. So I mean, so on the works. And we are only developing in Korea at the moment, but we're slowly trying to expand it. So that's what we're trying to achieve. I think it's the next generation education environment that we're trying to provide.
Alexander Sarlin:Yeah, that makes a lot of sense. It's really interesting. That also reminds me a little bit of Khan Academy has a lab school, you know, right by its offices, that is a full school, that's really excellent school, and they sort of use it to pilot some of the newest things in Khan Academy. It sounds like this QANDA Institute, that school is a lot like that. And you're thinking a lot about sort of classroom integration. It's such an amazing time for education, technology, you know, one of the things I wanted to ask you about, I have this recurring epiphany, I I'm like amazed, and then I forget it. And then I'm amazed, again, when you realize that that piece that you sort of mentioned in passing that language, the language barrier has really come down right pretty rapidly. You're incredibly successful, you know, South Korean company that was already crossing borders into a lot of Asian countries, and now is going around the world into Europe and the US. But that's potentially true of almost all education right now. And Korea, as you mentioned, is really famous for its private tutoring segment of its education. And there are these sort of super tutors in Korea, these people who are incredibly rich and famous, basically for being tutors. It's kind of blows my mind that we don't have anybody like that in the US. But do you envision people now that the language barrier is down and those teachers who teach even if they're teaching in the Korean language, you somebody could log in from Uganda? Or, you know, are France and watch them? Do you think that's going to change the landscape for global education and tutoring?
Kevin Leem:I think eventually, maybe. But I think in education, it's purely my point of view. I think language is definitely important, but it's one of the important thing that you can have on your list of the crucial elements of education. Because I think at the end of the day, what's important is knowing the local context, and having enough database to cater to each students and having the right teachers to actually help them study because no one can we really don't think AI can actually make someone study. So I think the mixture of that is something that's more crucial than the language barrier. So and nowadays on JD is you said the language barrier has been coming down and I think the game at the end of the day will be who knows the most local context and who has the most exclusive database to bring up that accurate slab on top of the accuracy that has has Built by this foundation models, the incredible ones. And the UI UX for students and the teachers to adopt and actually effectively use it for their studying method and who has that pool of human tutors can actually execute these by empowered by the AI technology tools that players like us are providing. So I think it has to be a mixture of all that rather than just one Gen AI the LLM foundation model the AGI, even the AGI, I think it will take time for them to come by even the AGI is there, but they will take time for them to actually motivate students in understand what the local context and provide the structure curriculum and education materials.
Alexander Sarlin:It's fun to hear you mentioned sort of AGI Yeah, AGI is coming in. But it'll take time, like the artificial general intelligence, which is this, you know, concept and AI. You know, I think our listeners mostly know about it. But it's interesting to hear you say with such confidence that it's sort of on its way, it's going to be a wild time. And I take your point, I mean, I think there is local context, it's not necessarily true that you know, somebody who is very good at, as you say, motivating and tutoring students in, you know, any one country might be ready, if you just sort of translated them instantly would be actually motivational or pedagogically nearly as sound in other contexts that is potentially true. My personal take is I'm really bullish on ideas like that, because, you know, we have this big world and education systems are so unequal, even within a country, as you mentioned, let alone between countries and between regions and between continents. So the idea of, you know, really, really high quality education being accessible everywhere is a dream for a lot of us in the tech industry. And I think AI, hopefully will, will speed that up in one way or another. Well, I'll have to see. Let's talk about the motivation piece really quickly. Because I think this is really key. You know, you mentioned that one of condors original core concepts was reducing the inequality between the sort of urban and rural regions in Korea. And one thing that a lot of students have in common is that they're all competitive, not all but many of them are competitive for spots in higher education or for jobs. And that competition serves as a as a sort of core motivator for standardized tests, does Kanda ever expect to move out of sort of the standardized testing world and try to teach things that that may be more, you know, based on learners interests, or something they want to learn outside of a formal testing context?
Kevin Leem:That's actually a very good question. And to be honest, I mean, this is something that we would love to address. At the end of the day, by just knowing how the education market, the standardized test market is the one of the biggest industry in the world that is still very, very fragmented. And there, students out there even in developed countries, that has not been receiving the enough equity education compared to their peers, so that they can not like climb the social ladder, there's so many students out there right now that we can be helpful by developing our tool. And we are nearly like 1/10 of it there. And I think that's the laser focus area that we want to focus right now. And clearly the areas that you just mentioned, it's will be definitely down our road by it's something that we haven't really fully given our thoughts through. As of now, we've
Alexander Sarlin:observed on the podcast that the Indian ad tech ecosystem, which has also, you know, which has really exploded in the last few years, similarly, many of the biggest companies are standardized test preparation in various ways. And it makes sense, because another very competitive country, where education has a huge impact on your sort of life prospects. And what you're saying makes a lot of sense. You know, if we're trying to build a sort of meritocratic society based on standardized testing, this type of educational technology that can help students get over that bar, and really improve their lives makes a lot of sense. Just sometimes I think some of the US edtech market really focuses on things that, you know, that are so far from standardized testing, that it's sometimes it feels like a schism. It's really, it's interesting to hear. I have one more question for you, before we sort of go into our final questions, which is, you know, what should people in the US really know about conda and Krama? Fi as you start to expand your presence here, like, what do you think are the core things we should be looking out for as more and more, you know, American students and teachers start saying, Oh, I'm using the conda app, and it's been amazing, like, what should they be looking for?
Kevin Leem:I think we really want to be remembered. And this something that really wants to build their content classifier is the best personalized AI tutor for everyone. So we really If the most effective education comes from this one to one tutoring provided by like high quality teachers, and that's something that we really want to build, so everyone can have it, regardless of their geographic or financial background. So what we have been focusing on UC Ponta and credit Fi is understanding who our users are, which school they're from, in which grade, and what they're studying for which classes which subjects and getting all that personalized information about them. And the study materials will really help them to prepare for what they need to prepare for, and shorten the length and the effort that you put on to maximize your learning outcome. I mean, there's a lot of debate about meritocracy. But I think there's a lot of benefits we want to give. And it's arguably true, that is something that a lot of students are in desperate need of, but they just don't get it. And that's something that we want to provide. And we just don't want to be in a position where just providing materials making them steady, steady, steady, but we want to make our AI tutor be for their friends. So if we know their I guess backgrounds and their schools and grades, you can be very hyper personalized. And we can provide not only the education materials, but whatever that is related to their school life, or like mentorship, and anything related that we can help them to guide their way, not only in their studying, but also like the careers and anything though. So we also have our community feature in our app, which helps like connect the students and share their feelings about what they're going through. And we're trying to connect them with the tutors on that front as well. So we want to do a lot more than making them like study. We know the Korea is really harsh in that, but we want to be bigger than that. But we want to start tackling by, I guess, providing what is really required for these students who are in desperate need of products like ours. That
Alexander Sarlin:makes a lot of sense. And I think the emphasis on the potential for personalization is really powerful and are clearly a major legacy of this AI era. This has been fascinating. Let's move on to our final questions, because I'm really curious to hear your answers to these, you know, the first is, what is the most exciting trend that you see in the education technology landscape right now that our listeners should keep an eye on. And, you know, this is particularly from your perspective as the CFO of Nespresso and cuando. Doing this global platform, what do you think is something that you see coming that others may not have seen yet?
Kevin Leem:I think it might sound like cliche, because so many people are talking about this ready, but I just can't stress enough, it's definitely generative AI. And not only because it's something that has been, I guess, mind blowing, and it's a game changer for a lot of things, but also as a something that we need to be cautious about about the hype. So I know AI and even the judge has talked about the problem solving ability is already already at a quite impressive level, and the link barriers has already coming down. But at the end of the day, the big tech players who are building this foundation model, they have to target massive audience because they are putting so much money on it. And they need the ROI on their investment. So they can cater to a specific group, let's say for like students in certain district or certain schools, or teachers, because all the feature has to cater to everyone. So I think this these are the niche areas that we should have our eyes on. And because Jenny is out there, I think it has changed two things in education. And I think whoever is in best position to execute this is will be the game changer in the tech industry. I think first is the content generation. And two is content delivery. So for the content generation, local context is so important, because there's no one size fit all curriculum, especially in K 12. And there's also reason why there's no single global leader in education. And I think that's similar to your point that you're brought, I think it will start become possible because we can create content at scale level regardless of language, which can cater to each region at very low cost and high speed. And as long as you have the relevant data. So that comes the importance of data part that I explained earlier. And the second part is the content delivery. And I think this teacher shortage issue is a global phenomenon. I haven't seen a lot of news about this, especially in the US after COVID. And I think Gen AI is in the best position to solve this issue by utilizing AI tutors and the AI TAS which would help teacher to focus on more important aspects of education. So now that because these two problems as splain is now solvable, which an AI not right now at the moment because the cost of Gen AI is so high, but eventually once the hype is gone, I think there will be confused players who will be able to execute this and we're just hoping the content will be one of them. That's
Alexander Sarlin:very, very well said. And I think your point about the big tech companies that are doing the foundation models, the the Geminis, the chat up to the cloud, you know, Sonic world that we're seeing right now, you're right there for everybody. There are a global audience, they're not going to be tailored specifically to education. I've actually been surprised at how much they have paid attention to education this early on. But yes, I agree with you that the edtech companies that are going to succeed are the ones that will be able to sort of harness that incredible power that's going to be you know, continuously growing for the education context. And I think your emphasis on localization is very, very interesting. I, it gives me a lot of food for thought about global ed tech solutions or things that can go across Europe, or, you know, we talk a lot about that on the show, but I think I'm gonna really mole on this idea of you know, how localization and education really might be very contextual in different regions. So our final final question is, what is a resource that you would recommend for somebody who wants to dive deeper into any of the topics we discussed today? This could be a book, blog, newsletter or Twitter feed or your ex feed, you know, anything that will help people go deeper? If they are finding this conversation really interested in they want to know more is
Kevin Leem:other than the tech insiders? Yes, yes. In that case, I mean, I read a lot of articles and books, because especially there's so many things changing in this world right now. But I would say one of the things that I am, I guess, really excited to receive in our in my inbox is a things GSV newsletter on AI. And I found them very, very helpful, because I'm based in Seoul, Korea. And it's very fascinating to hear all the insights from these tech investor from Silicon Valley on how this texting is evolving so rapidly. I mean, it's it's crazy on how like a lot of the tech players are adopting Gen AI and everyone just talking about the Gen AI and just knowing what is going on, and how not only the tech players, but all the tech players are trying to use this technology to make their product better. I think it's interesting to hear this straight from the team there. And I guess the last thing I just want to know is I attended also the ASU GSB event they held earlier this year, and was mind blowing to see a lot of the AI related products popping up. And I think at the same time, what I felt was we need to be very cautious about the hype. But I am really excited to see how things will change even more by this time next year, including our own products conda and credit fi and how they could really impact students learning behaviors with technology. Yes, that's that's one of the main source that I'm looking forward to.
Alexander Sarlin:Yeah, absolutely. And that's Claire's ours, amazing AI newsletter GSV. Must Read. We love Claire on the show, which she is going to be a guest host for some of our episodes soon, because she was just so knowledgeable, as you said, and they are so knowledgeable. And congratulations to her. She was just became a partner at GSP. So that's big news through well deserved and focusing directly on AI and education. So we're gonna always hear more from her. She's absolutely amazing. As always, we will put links to the resources in the episode. This has been so interesting, and I really appreciate it. I feel like I've heard about mas reso. And on the sort of as a, you know, on the side, knowing that it's a big thing in Korea and in Asia, but hearing the actual story behind it, hearing the equity vision for it and how it's been doing AI for almost 10 years and is writing this generative AI wave in very, very interesting ways. Just gives me and I imagine many of our listeners a whole new perspective on it and the potential so we really appreciate it. This is Kevin Leem, the CFO of Mathpresso operator of the globally recognized AI powered learning platform QANDA, which serves over 90 million users and I'm sure growing everyday. Thank you so much for being here with us, Kevin on a tech insiders.
Kevin Leem:Thanks, Alex was a pleasure.
Alexander Sarlin:Thanks for listening to this episode of EdTech Insiders. If you liked the podcast, remember to rate it and share it with others in the EdTech community. For those who want even more EdTech Insider subscribe to the free Edtech Insiders newsletter on substack.
