How Do We Teach 2B Children? Democratizing Access to Tutors With AI | Nhon Ma of Numerade

Listen on
November 11, 2021

Nhon Ma is the CEO and Co-Founder of Numerade, a Los Angeles based EdTech platform that’s using AI and short form videos to democratize access to tutoring for kids around the world. What is brilliant about their idea is that they are bringing supplemental learning to students who can’t afford traditional private tutoring.

Nhon is an ex-Googler so he’s a tech-savvy founder who just knows how to run a business.

During the chat, Nhon shared his extremely inspiring personal journey and if you follow the thread of his story it seems as though he was destined to start this company. We discussed the intricacies of Numerade, how learning varies by states and countries, how education is evolving in the US and so much more.

Show Links

Transcript

MPD: Welcome Nhon. Thanks for being on today. Appreciate it. So let's start off. Do you mind giving just a quick overview of your background?

Nhon Ma: Yeah, definitely. I grew up in and was raised in LA. Went to school here through public and then private school was fortunate to go into private school here in LA, then ultimately went out to the east coast and went to Columbia for four years there, majoring in economics and psychology.

Shortly after that, did the whole two years in finance realize that, Hey, that wasn't it. Then moved back out to the west coast and the bay area worked at Google for a while. A collective seven years. In between that actually left after four years. This is after stints in sales strategy, analytics and operations did my first startup out in New York.

It was. We were the anti Groupon at that time, and then be able to be pivot towards a social music site, unfortunately, that didn't really get to a point where we can scale the business. So we ended up selling that off I boomeranged back into Google where I was a, the product lead for their digital market.

A platform, the programmatic ad business. And during that time was working on my, by my side business, which was in ed tech. And that eventually led to what today, which is Numerade, super quick, high level.

Okay. Couple of questions for you. First. What town are you from? An LA. I was born and raised in west Covina.From out there,

Nhon Ma: We live in Pasadena now, but I grew up in south central LA, so east of the one, 10 freeway. And so Jeff in the area grew up in the eighties and nineties.

MPD: That's got a tough rap. Was that a difficult childhood for you or

Nhon Ma: definitely wasn't easy. I would say that, my parents came over as Vietnamese refugees with were Chinese Vietnamese refugees.

They were the second wave of migration. So the Chinese had left Vietnam, escaped communism, my family, unfortunately, all their property was taken away. And so they settled in Houston, but they shortly moved over to LA. And at that point, we didn't have any, anything at all. You know, the most affordable place in LA was south central LA.

I think growing up it was definitely hard in the sense that we were the only Chinese Asians in that area. So a lot of it was my brother and I just had to really fend for ourselves to figure out, how, how we can ask her to survive. But ultimately um, both of us, kind of coming together partnering in a way that made it work in a sense, but definitely not easy, not easy.

MPD: Must've been a hell of a culture shock showing at the Columbia

Nhon Ma: university. I got to say it's. So I went through the LA USD school system. You know, from. Kindergarten through to junior high. So this was in the eighties and early nineties. So everything that you know of south central LA, was clear.

But actually my biggest culture shock was actually in high school. I was switching it to get a scholarship through a program called a better chance that allows for, inner city kids to get into these college preparatory schools. And so I was able to get a scholarship into a school called Campbell Hall.

It was an amazing school, amazing friends and connections there. But my biggest culture shock was there going from south central LA to a school in the valley where now tuition is near that of an Ivy league.

MPD: Wow. Okay. Got it. So you had already been acclimated a little bit from high school and

definitely.

Nhon Ma: Now I say going into Columbia was a different type of culture shock. I think level of, of of students there and just the, you know, the caliber was, is at a different level. So I'd say it's both were definitely challenging and culture shock, but I think the early days deal with that sense of what we have to do to survive in these certain situations. And so I think that was super helpful.

MPD: I love the self-made story because it's so much more fun to root for folks in that situation. And I know you're out doing great things now, which we're going to get into.

So kudos on that. What you want to start off by telling us about Numerade? Cause I know you're off to the races with this one. It's a big concept. I think people would love to hear this.

Nhon Ma: Yeah. So we've been working on Numerade for the past two and a half years, but it's actually been a seven, eight year journey to get to this point.

It doesn't, back in the day when I was at, which was a telomere of my career at Google, we were starting to dabble it and think through really a larger concept within education and, you know, given the background I had it was clear to myself, my co-founder Alex that there, there are.

Discrepancy of access towards educational opportunity. And so when we came together, we created a different company seven years ago called tutor casts that Ashley evolved into Numerade. And so what we wanted to do initially was to decrease the costs of tutoring, right? So a lot of my peers at my private school had tutors.

I would go home right in that wasn't the case without peers I grew up with. And we want it to decrease the cost of tutoring by building this platform for synchronous tutoring now in a post pandemic world and what that's kind the norm now, right? Being able to go online with distance learning.

But back then, this was still new. And so we wanted to decrease the costs of tutoring using the synchronous model when we, so we sold it. It's a lot of independent tutors, tutoring agencies, and even communities. The reality was it's really hard to decrease the cost of tutoring using a synchronous model because you're dealing with a finite constraint that find a constraint is a tutor's time.

So a great tutor in the upper east side of New York who's tutoring sat will continue to charge 300, 400, $500 an hour for her tutor. So there has to be a different model. And what we realized was that using a short form video approach can actually work because when we looked at tutor casts, one of the uniqueness of tutor casts was that you can actually record all the sessions.

And so interesting behavior, MERS students are going back into their replays to review the content over and over again. And when we did quality assurance on the, these videos, we realized that tutors are providing these explanations on how to approach certain problems. So the content was interesting and also at the behavior of the students was one such that, they were actually learning deeper by going back and reviewing content in their own space and time, but also getting the reps in, especially with stem education you had to get the reps in to learn the material. And so that was a big light bulb moment for us. We realized, what do we can take that content and scale it out to the masses? And perhaps the benefits of tutoring can be scaled out to the masses.

And so that evolved to Numerade where, what we realize, what we're doing is that we're looking now to imagine if you can digitize and record the knowledge of every single educator. Or student or anyone who has some mastery over any subject subtopic or even way of explaining a problem, record that, and then make it universally accessible to as many students as possible in a very personalized way.

And so new raid was, has that foundation of what we're trying to do. And I would say, as we moved from, the original mission to low the cost of. We want every single stew in the world. It's not even half the pay anymore for, a 50, $40, even $20 an hour for a per hour for a tutor.

They can actually reap the benefits of tutoring through our platform. And so when we think about what we're doing we're in a larger mission now because we realize what we can do, which is to support students at an earlier age to graduate more students into the stem field. And, and the reason why you want to do this, if you look at the past, few, few months or 12 months or so, and more, it's amazing how we've been able to get to a vaccine in 12 months, right?

There's so many more problems and issues facing humanity. And our fundamental belief is that graduating more folks into the stem fields will help alleviate a lot of these problems more and the rate of progress within humanity would just be a celebrated at a much faster. And so we want to start young with middle school students for now where the student who is in south central LA or east of lake, who doesn't have access with tutor is able to get the support.

They need to understand pre-algebra geometry trigonometry. And continue to get the support they need, because we know within the, within this period, oftentimes if you don't get the support, you need, you fall off the stem trap. So we want to continue to provide that reinforcement to the student support throughout their academic career in, uh, in, in stem.

And so essentially what we're building is an AI tutor that leverages all this type of content.

So that that mission, obviously the broader mission. As, resonates with us, what we're trying to do, what we hope to do is play a role in helping to accelerate innovation as well.

I think it's probably a mission that a lot of people in the entrepreneurial ecosystem share but the more narrow tactical approach here of doing that by getting delivering access to tutoring, to people who maybe can't otherwise afford it. Does that come from your personal background? The experiences you were telling, it sounds like it evolved the way you tell it out of the work you were doing at Google and what you saw and what you learned.

But having, we just went down that rabbit hole on your story a minute ago, when you put those two things together, they do seem to fit pretty well. So is this a little bit of a personal mission as well? Or is it just merely the opportunity you saw at Google? And it happens to sound like it might be a personal mission,

right?

I think that is as an entrepreneur. I think that's critical to continuing forward with what you need to do to get to the next step. But this is definitely a personal mission. I would say, definitely getting some eye into this into this high school when getting into Columbia, a lot, a lot of folks did have these access points to tutoring.

And then I see so many other folks in various areas within LA, throughout throughout the world who, who lack that. Particularly with the tutoring to begin with. And so it is definitely a personal mission, but I think with, working at Google was a, it was a hugely insightful and transformative experience for me, where it gave me the insight of how technology can play in in, in making information just accessible to as many people as possible.

And um, I think those two were, were key factors to the, how, how, how, how new radio.

MPD: Have you seen, I know I'm going to get a little bit more into the company, but while we're on this thread of social impact, have you seen any patterns in usage or acknowledgement from people maybe who couldn't afford tutoring otherwise?

Do you know that it's reaching the folks you want it to

Nhon Ma: most? Yeah, definitely. You know, we've have over 20 million students have used our products And I would say that the feedback we've gotten has been overwhelmingly positive in the sense that it's all strata of socioeconomic backgrounds.

So we have students, in, in different areas within really the world, be able to access this, this, this continent service. And we're hearing this directly from students saying that we've been able to. Help them get into the college. They want help them get the a in the physics course where they couldn't afford tutoring.

And so, uh, and as I say, those are the stories that help us continue on this mission of continuing to provide the best product and service to these folks.

MPD: Those are really powerful anecdotes. As a founder, we recently launched a Social good service, a company called thunder.vc, which matches entrepreneurs, angels, and VCs.

And just this week, someone emailed us and said, they found that told us they found their lead investor on it. That's awesome. It is super uplifting when you put yourself out there to build something like this and you hear that people are getting value from it. So kudos to you. Are you putting those testimonials on the homepage and all that?

Or are they

Nhon Ma: some, yeah.

MPD: I bet you there's a lot though at 20

Nhon Ma: million people a lot. It's way too many right now, but definitely some great feedback.

MPD: That's awesome. So can you take us through how this platform is different than some other solutions out there? There's a number of strong players in the digitizing education space.

How do you think about, your positioning and how it's different and how it's complimentary to what people do.

Nhon Ma: That's the Mo when we think of Numerade and the way students interact with Numerade is this I think the the key metric to talk through is your average video length is acid.

There is short it's about two minutes, so it fits the gen Z consumption patterns. So you got a two minute video on average, but your consumption times are six to seven minutes where students are rewatching the videos over and over again, get into specific moments in that video. But when we think about, how this works and the larger notion of, of, of how we differ, w we, we are creating the AI tutor, system and, and really, overall service out there that, that brings in a human elements, who it, when we think of what's happening out there in the AI tutor landscape these are, there's a lot of calculators out there that, you could take a picture of something and then you get a step-by-step in an answer.

We think that's fine. That's okay. But that doesn't really help drive. Deep learning comes from sight sound narrative. It comes from the human element of learning. My mom was a teacher back in Vietnam. My dad was an entrepreneur in Vietnam as well. And so it, it's interesting. I'm in the ed tech space.

It's all coming together. But when we think about teaching, it requires. Person to deliver that to the student, my son goes to the math naseum and, and, there's a specific way the teacher there teaches how to add and solve. And so it's hard to capture that right now from a strictly, uh, kind of a calculator type of approach.

And so our biggest differentiator is the fact that we tap into humans to then articulate how to solve particular problems and shine a light on the. So we have over 1.5 million videos, um, you know, that's looking to expand in a really big way. And so starting with video is really the biggest differentiator.

And so as a student, they'll come to us through an inquiry. And so oftentimes you'll see students. They have, they have heart, they have questions, we'll always have questions then stem. When I was, take my cow cores, chemistry, physics, you always had questions. And so it's really hard for a student in a classroom environment to raise their hands.

If you don't want to look stupid. And so oftentimes they would go online to find resources and every single time they do that, that's a, that's a potential inquiry that's hell's us, that students, they still need help with certain items. And so they'll come to us, you'll see other competitors, but when they see us.

They, they they're they're, they're greeted by an actual educator that has provided this a custom video that allows them to deeply understand how to approach a problem. And so as a student goes on to the site, And know back in Google, I was a product lead for the digital marketing you know, platform there that the programmatic ads business and when the key technology is there and it's going to be no longer the case soon.

And it could cooker this world. What is the concept of cookie where, you know, everywhere you go on the internet, you, the cookies, your profile of your behavior, and you get targeted as based on that, right? For us, we leverage the student profile in a very similar context. So instead of pushing consumerism of stuff, that's folks may not need, we're building out a student profile, which is a detailed accounting of what a student is like.

Where their gaps are in their knowledge as insured by their watch history. So essentially we are building the tick-tock for education. So whenever a student go comes to the site as they go on to other sites, we know the next sequence of, of kind of problems, concepts, and we're looking to identify gaps in their knowledge as well, based on their watch history.

So then go back to say, okay, this piece within geometry, you're not clear on, so let's go back to that. So the platform actually has over a 1.5 ish million videos that we call practical videos, which are a video tutorials, a specific problems, and then there's, we have 17 courses of actual stem content, right?

So we have this huge educator base of over 35,000 educators. Roughly 70% of them are actual teachers. And, and a, and a percentage of them are actual PhDs currently TA uh, within, top colleges. And so for example, one of our, one of our chemistry, one of our chemistry courses is, this, uh, this TA who's pursuing her chemistry PhD in Caltech, and she has a whole 30 hour um, series of a whole entire chemistry course.

On the platform. So for us, the AI tutor is essentially a recommendation system. Now that shows the user based on how we infer, with our learning, with our weaknesses, our concept videos, practical videos, and also an AI quiz. That's all awesome.

MPD: And I know the mission is to give access to this educational training and believe you guys had just launched a program.

Around Afghanistan. Yes. You wanna explain that?

Nhon Ma: Yes. So we're, we have a program now where you refugees, uh, from Afghanistan who, you know who are in stem or learning stem can get Numeraded for free. I think that this that whole entire story resonates well, I have a picture of here uh, of my mom's.

So she was on a boat as similar, a similar type of story. I think that imagery resonates very well. Where her students, she was a teacher or student Ashley painted this picture of her on a boat with all the other folks leaving Vietnam. And so when we saw what happened there without Ghana, Stan, we realized, Hey, we had a, we have a huge responsibility to help out as much as we can with our platform.

And so any student that needs help in stem, uh, can access the Cyprus.

MPD: Wow. That's fantastic. Now, when you're serving different countries and different cultures within the United States, beyond the United States, is do you have to tailor or think about the algorithm and the product differently?

I'd imagine people generally learn the information in the same sequence, but that's not even probably a given. And I know the states doesn't have the highest ranked educational program. Some other countries might move faster. How do you think about. Adapting totally

Nhon Ma: product. Totally. You know, every single call, every single with our state does have its own unique curriculum.

And so the way we're adapting is really through our educator base and also how we cure. The curriculum by, so say for example, Texas may have, it's all a form of a poor curriculum relative to, uh, California. And so we do have a system by which we can cure, rate the content into individual playlist that is specific to the, the standards of, uh, uh, Uh, of a state.

And so, and, and I would say the other piece too, is because we have educators from across America and around the world, they also bring about getting back to the human element of learning those, uh, those standards and ways of teaching that resonate well with that local community.

MPD: Got it. So you've got just enough volume of videos to adapt for different geographies.

Maybe different cultural needs. And are you doing different language

Nhon Ma: as well? That's that's definitely on the roadmap right now is all English. Okay.

MPD: Now let's talk about education because you're obviously deep in this. It feels like your whole personal story has led to this moment, by the way, once you shared your parents' backgrounds as well, it just feels like all roads for a couple of generations have been building to here.

How has the education you're closer to it that I am changing in the U S we just went through a pandemic. I know that had great effect. W what is happening in the U S market? Again? I know you're focused in the business side on tutoring but I'm interested to hear your broader perspective on education and how it's evolving in the us.

Nhon Ma: I think there's a short term evolution. When there's a larger responsibility that we have as an organization. But I think there's also a mid and longer term impact of COVID. Clearly the adoption is the biggest one, right? So you have less of you know, what would COVID has provided especially for parents is a window into what is happening in the class.

And I think what is clear is that. Depending on where you are you know, this is a case for me with my first grader. You see what was happening. You did see teaching towards the mean because you had it. It was just it was just the case. It was a one to many type of teaching situations. You know, the current system or of education has been outdated for the past.

If the plus a hundred years, it's a one to many model, that model I think COVID has. To say that it doesn't make any sense at all. There's certain students that need more help. There's other students that don't need that much help. And they actually need to be a celebrated on a different track.

And, and, and the current system doesn't really fit that mode. And so going forward, fast forward, it's happening now, you're seeing more parents take their students out of public school and homeschooling. Um, you're seeing a lot of that happen and what we believe what's going to happen the next, five, 10 years is a celebration of of choice for parents to understand and figure out, okay.

W where, you who, who, who, you know, what's the best educator for their student? W what's the type of content in which they can, they can receive. But we believe that everything will be personalized in a way where every single student who wants to learn or needs to learn, say pre-algebra calculus to trigonometry and get that in a way that is highly personalized to them in these Mo in these modules.

So think of a situation where a student in south central LA, who doesn't have access to an AP. Computer science teacher, right? Can you hand get that instruction? That's highly personalized to that student in and get the credits they need as well. Now, granted there's issues inherent in a lot of that too, with the digital divide and how much you can get from a harvest and wifi standpoint, but we believe the future of education is one that's going, gonna be highly personalized.

Content we'll still be king, but it's really the best content. Being able to tap into the minds and knowledge of the best educators, but also the best students, that I've actually had some insight into chicken arbitrary, capture that in a way that, that resonates with the student to continue that learning process over and over again is what we see the future to me.

MPD: That's interesting. So if you went internationally, How does that change your business? You're based in LA your contents in English. Obviously the language has to change. What else happens? It's different for entrepreneurs. This is, I'm asking this less for you. And more as a case study for entrepreneurs listening, we're considering an international play.

What do you need to do to be successful going

Nhon Ma: abroad? Yeah, I think what's very important and this kind of gets back to. What's needed to be successful to even begin with. I think it's, I think it's team I think having a localized presence is very critical.

MPD: Why, why do you need local

Nhon Ma: people?

There is these cultural nuances that is is hard to understand if you're not there. So taking the case of China, there's clearly going to be a lot of regulation. You know, the there's different curriculum. There's probably different go-to-market strategies to write. And in Europe, Southeast Asia, very similar, right?

The type of things that's, students will be learning different systems are going to be very different. And so having that localized presence is key, but what's even more key is finding you know, there's this interesting concept. Uh, keeper boy, who was a big investor, talks about a lot, which is the concept barrels versus ammunition.

You want to hire barrels in where these 10 X-ers those in which you bring on board will, can be leaders, can, we can drive market expansion into these different areas. And so I think key is you have to find great people who can be leaders in these regions and take it to the next level. I think through all those cultural nuances, that's what.

Okay.

MPD: So the, the main tip of advice for folks would be hire local, expand your team locally. How do you do that? Right. Even, managing offshore teams, we have a couple of businesses with international footprints. There's cultural differences, that management phase there's communication style, there's different values in some areas, right?

Some countries are less progress. How do you any tips for navigating that or preparing for that at the very least?

Nhon Ma: I think it's tough. I do think it starts with finding the right set of folks to be your leaders in market. You have to be, first of all, finding them is the hardest thing, so finding them first, that's step number wide and to ensure, during the interview process you're aligned in terms of value structure, mission, and culture.

So it has to start there on, when, when folks come in. And then, and then afterwards, this is this an operating cadence in which you have to ensure that you're aligned on a day-to-day basis against your goals. And, and our view, we operate, and this is really from working at Google, being very metrics and data focus.

We, we're pretty obsessed with data. And if, if, um, you have to build an infrastructure to manage your operation, especially international and using data as a way to understand the ebbs and flows of the market to give you a level of comfort and also a way to figure out what needs to be focused on internationally is critical too.

So your data infrastructure is going to be key.

MPD: The hiring tactics, you've seen something as simple as. Yeah, hire people who lived in the states for awhile. So you have some commonplace at least on the U S culture or is that not the trick, but what do you look for?

Nhon Ma: I would say a lot of his network and, so we have success with looking within our network to see who are these great candidates that we can we trust, who knows how to do it?

You know, we haven't really done much in terms of, the, the normal routes of, putting, putting. A job rec out yet on LinkedIn and such, we're going to experiment, but a lot of it is network right now.

MPD: I guess what I'm asking is what do you look for when you pick somebody?

Oh, gotcha. For to be a bridge, to.

Nhon Ma: I think experience doing this before is probably the biggest one. So someone who's actually done this work either consulting or working with a US-based company and has already done it one or two, three times before. I know these are rare folks to find, but if you find a jump on them and you know, put them on, put them onto your team base app.

MPD: So taking all of this in you have great perspective. On seeing different educational patterns talked a little bit about what you're seeing in the pandemic. You're looking into the classroom and seeing how people are teaching to the mean that's not good for any of the extremes. If you were king, how would you change the U S what policies would you put in place?

How would we go out and improve our educational system? I feel like most of us are here that were bad in the news. I don't know what, if anything we could be doing. Yeah.

Nhon Ma: There, I think there's a component of learning that if I were the president and had some real authority to affect real change in a way for education, the approach that I would take, there's really two fold.

There's one of just straight up learning which. You mine are our fundamental belief is one where you have a system that is driven by personalization, where the best teachers and educators who have the best mastery of the material, but the way in which they convey the material resonates well, they can be that can be a one to many type of situation, where they are. They're either live streaming. In, in a broad capacity, because you want to leverage the best organic chemistry teacher and push that to as many students as possible. At the same time, there also needs to be a TA type of system where there's teacher. But they not, they may not be the most charismatic right.

But they know the information inside and out, they're there to help us very much like a tutoring situation. So it's essentially live streaming of the best educator and individualized tutoring in the school. Where, um, you have these instant feedback loops that's built in to their consumption of the content and the TA has.

Inputs by which the TA can then say, all uh, let me assess this from a a student by student basis, right? So you have this notion of let's scale out the best teachers has many students as possible, so that you don't have kind of this unequal distribution of best teachers reside in certain schools.

But I also, at the same time, let's still cultivate these educators that we have and allow them to basically help their students in an individualized fashion. Where technology is really the catalyst that brings everything together from an assessment standpoint. That's how

MPD: didn't Khan academy.

Try to do