Driving Educational Outcomes Through Evidence-Based Learning and AI

00:00.00

Dipesh

Hello everyone! Welcome to another episode of Tech in EdTech, a show where we talk about technology, education, and everything in between. Our guest for today is Sunil Gunderia. Sunil is the Chief Innovation Officer at Age of Learning, well-known for ABCmouse, the creators of ABCmouse, and many other products. I'm very excited to have Sunil on the show because of 3 things - early learning, evidence, and AI, and we will dive into all of these 3 topics. Thank you, Sunil, for joining the show. Welcome to the show.

00:38.40

Sunil Gunderia

Dipesh, thank you for having us! I'm really excited to be here to talk about what we do at Age of Learning, and the opportunity we have to really make a difference in the lives of children and students everywhere.

00:50.80

Dipesh

Thank you, and Sunil I mean I'm connected with you on Linkedin. So when I looked at your profile, I loved that you spoke so much about evidence because I think that's a gap that I see in the current industries. So we'll definitely dive into evidence. We will of course dive into AI. That's one thing that everyone is talking about, everyone wants to learn more about. And I'm personally very excited because I have two kids, a 5-year-old and a 1-year-old, and early childhood education and how AI intersects with that is going to be very exciting for me and I'm sure a lot of our audience. Before we dive into all of these meaty topics, I would love for you to give a little bit of background about yourself. How you got to where you got to and talk about your journey in edtech.

01:32.80

Sunil Gunderia

Absolutely, happy to share my background. So I got here in an indirect way, in that I started my career in finance and accounting, really following in the footsteps of my father - I'm a first-generation immigrant - here to the US from India. And you know, I didn't know better when I was growing up of all the opportunities that existed with entrepreneurship and an opportunity to really make a difference and do meaningful work. I really discovered that first, the opportunity in technology working for Disney, which I did for 11 years. Both here in the US and in Europe, where I was responsible for working on new mobile platforms like devices and bringing Disney content there. I left Disney and saw the opportunity, and how kids were really engaged with content on mobile devices and the interface worked really well. And I was looking for opportunities with technology and education and was fortunate to find Age of Learning which had launched ABCmouse. I was the first executive hire to the Age of Learning team that was not on the founding team and I was brought in because of my mobile expertise, my international expertise, and my understanding of mobile interfaces. I quickly took on the Chief Strategy Officer role and then to my current role where I'm now as Chief Innovation Officer I spent 11 years at the company, and have been really fortunate to drive the development and distribution of products that are really impactful in the lives of students. And you mentioned outcomes in efficacy and we have a lot of evidence to tell you that our products are going to drive favorable outcomes for kids. Both our legacy product that I believe you have used with your child - ABCmouse as well as our new products that we've developed for the classroom - My Math Academy and My Reading Academy.

03:50.36

Dipesh

And that's great! Exciting journey Sunil, I mean it's very interesting to see how from Disney to education and how that, you know what you call it the core concepts of gamifying - How do you bring it to the educational ecosystem? And I think that's very exciting. So I have 2 questions for you and you can answer in the order that you'd like to. One is a burning question for a lot of us is how do you - there's an emergence of AI, definitely, AI is on the rise - but when you combine it with early childhood learning or early education - how do you merge both of those? I mean there's a lot of applications of AI as you grow up. But when you see, when you talk about kids of a young age group, how do you do that? That's question number one. Question number two is - which is personally the reason why I chose ABCmouse - is your focus on evidence and impact, because I think parents are very weary about screen time so you want to choose products that you know are going to be great for your kids and they have evidence attached. So how do you merge this evidence and how do you measure evidence of AI in the learning product? I hope you got both the questions. Would you like to answer first, second or whichever that you prefer?

05:10.57

Sunil Gunderia

Let me start with the questions around ABCmouse and that is a platform that we have developed. And really because it is the story of the company is - through ABCmouse what we learned about is how to create a product that both meets the needs of parents and the needs of students. And the way we've done that is by making learning - which it is - very engaging for the effective child and through the results that we can get for students that have been proven by evidence, for parents. And really that allowed us to build a lot of trust early on in terms of how do we merge the joy of learning with a digital product and gain the trust of parents through that. And we invested in creating something that you know you've seen with your own child that they really love to do and as we all know learning is quite fun. And unfortunately what ends up happening is that when we go to school, because of certain measurements that are put around it we take a lot of the fun out of learning inadvertently, which is really natural and intrinsic to us. And in terms of the question around AI what is essential to develop effective systems that use AI - are domain data. Specifically on what a student is doing. What is motivating their learning? And how are we measuring the impact of our solution? And you'll hear this talked about a lot with AI - it's more than technology. It is not simply the Large Language Model alone. Its Large Language Model combined with domain expertise and data on what is actually happening during a process. And it's specific to our process is - what is happening during learning? And that is something that we've really built upon. And our research bears it out in terms of ABCmouse in terms of how effective a product it has been and continues to be for parents as well as the young learners that use it. We then applied what we learned from ABCmouse when we developed our products for school - My Math Academy and My Reading Academy knowing that collecting the data on reading and math and how that happens. And what's required to get a child from no knowledge in the subject area like reading and math to a skill like being able to read fluently or in math - being able to count at a second-grade level, which is where our products go through right now. It is the ability to add and subtract 3-digit numbers is where we get to as our penultimate skill for math. We have to then backward design every single element of learning into a knowledge map that really provides a scope and sequence of knowledge that a child needs to acquire as they go from no skill to our penultimate skills in both reading and math. And then creating a data structure that allows us to capture every click that a child is making in that journey as well as being able to measure what they're learning through things like evidence-centered design. So we've designed our products, My Math Academy and My Reading Academy not only to collect clicks but actually to collect what's happening inside a child's head, which is what are they learning and how do we know they're learning. And that was really essential into our design and what that enables us to do is now as AI features become more accessible; we're able to connect that data to not only make the learning better for the child but also to inform the teacher and the parent with what's happening in real-time in terms of learning for each student.

09:42.24

Dipesh

That's amazing. What you just told is actually kind of, I mean it's very interesting your journey from a B2C product to a B2B product. And right now you probably laid out the groundwork you did to make that journey possible. From being a B2C product getting the knowledge and experience that you got from there and applying that to a B2B which is a more school-based product and I think that's very interesting. Do you want to talk a little bit more about how you went through that journey from ABCmouse to My Reading Academy and My Math Academy? How was the journey for you personally?

10:20.36

Sunil Gunderia

The journey and it was a team journey led by an underlying truth that is very unfortunate for us in the US and it exists in countries across the world is that our systems currently are not getting the majority of our students to proficiency in math and reading and you look at the National Assessment for Educational Progress (NAEP), the NAEP scores - which is known as our national report card. Fourth-grade students, over sixty percent of our fourth-grade students, are not proficient in math and reading. 

10:57.36

Dipesh

Wow.

11:16.20

Sunil Gunderia

And if you'll look at what it is for our low-income students, our low-income students represent 50% of our kids in public education in the US. Nearly eighty percent of those students are not proficient in math or reading and the impact on students' lives, the lives of those in their community as well as the broader economic impact for our country is pretty severe, especially in a world that is being changed so rapidly by technology and AI. And that's really the origin story for My Math and My Reading Academy and why we started the research and development process back in 2015. We knew you know, with the success we had built in ABCmouse and the understanding of students, that we could work together if we brought together the right team of experts to really tackle and solve this problem. And that is really the origin story for why we're doing what we're doing. And to do that, let me talk a little bit about the team that it took to put that together. I talked a little bit about the idea of knowledge maps and backward designing knowing we had endpoints. We wanted to get a child through second grade, which was you know the ability to read proficiently. And for math, to be able to add and subtract 3-digit numbers. So we needed master educators that really could unpack standards and understood all the steps that it takes to get a child from no knowledge to those endpoints. We needed design research because one of the things that we learned through this process is that although all of us have been young at one point, we are no longer young. We're certainly not four or five years olds and we needed to bring them in to test that our interfaces work for them, that they were developmentally appropriate. And nothing better as a test of truth than to get feedback from a 5-year-old who tells you exactly what's on their mind in terms of your product and so that was a really important part of the product. And then understanding through the learning sciences and this is really fundamental. You mentioned game-based systems which we've used, what we wanted to do there was to ensure that we were applying, you know, the best in behavioral science in terms of how we motivate learning. The best of the learning sciences or cognitive sciences in terms of how we as humans learn and our memory works and the third pillar of this in terms of the learning sciences was learning analytics. How do we measure learning? How do we know learning is happening? And what are our bases for that? So I'm going to point out a couple of researchers that really are a few researchers that really informed our work. There’s Benjamin Bloom who talks about the 2 sigma problem and effectively to simplify, it is basically that a mastery-based learning system using a one-to-one tutor is the most effective way to learn and actually if you can do that it has 2 standard deviations better results than your standard teaching methodologies. So you know goal 1 of ours was how do we replicate that with the technology platform knowing that one-to-one tutoring is difficult to scale. Number 2 would have been Vygotsky, and really understanding from a motivation perspective placing a child in front of what they're ready to learn next is really important. And what that requires is you have to understand what a child already knows and is ready to learn next and then present that content. So that's exactly what our system does with the knowledge map. We start with an assessment that determines where a child is in their learning journey and then we serve them up the instructional content they need next. And the third researcher I would point out is Bob Misleavy. And his work on Evidence-Centered Design motivated us to look at how we determine how learning actually happens. So effectively the games in our programs act as assessments so that we know that a child has learned or not learned something before we move them forward. So we can show mastery just as I mentioned with Bloom and part of his system that we can show that so that each child has their own individual learning trajectory based on the knowledge they're gaining through our system.

15:56.74

Dipesh

Wow, that's amazing you know. I mean just a note for the audience. It just goes to show how much and what all goes into creating these learning products that students use on a daily basis. And it's amazing to see the amount of research and evidence and the right kind of people that you've taken to get to this journey. One thing that I noted here and also in what you've said I think we talk about students, we talk about teachers, and rightly so. We don't talk a lot about parents. In one of your episodes Sunil, I remember you mentioned that your product sends notifications or nudges that your kid has learned to count up to 20. I mean as a parent that's such great news for me. I would like to…if you could elaborate a little bit more about how you support parents and teachers and how you enable them. I think our audience would really like to hear that.

16:58.81

Sunil Gunderia

So the way we do this is the system that I describe to you is something we've patented. It's called our Personalized Mastery Learning System and as we explored that system we and thought about Bloom’s 2 sigma and the objects of change and Bloom describes objects of change and those things that have effect on learning and one of those systems are the actual learning materials. And the students themselves and how they're thinking about their learning. And then the other two are your home environment and your teachers and so based on that we developed something that we've written about several times called a Personalized Mastery Learning Ecosystem. So as a student interacts with our program, we are recording as I strugglingmentioned every click. Approximately in a classroom, a hundred clicks every minute are being met, we're getting telemetry from a classroom. And this is that at the individual student level, we are now able to track in the classroom for the case of the teacher which is, you know, a very important part of this ecosystem of learning. We are able to provide them real-time data of not only what a child is working on but also if they're struggling. If they're moving forward as needed. And of course, if they're struggling we suggest that a teacher may want to intervene and we actually even provide the teacher material for them to use with a student that might be struggling. If a student is doing well, we provide them material on - the teacher material on - how to transfer that knowledge, the things that they've learned on our platform, to other scenarios that might exist; so that they can see that the applied understanding of that knowledge exists. As you think about the parent piece of this, it was really important for us. The parent tends to be the most underutilized resource in learning and a lot of times it's because parents themselves worry that they're going to mess up what the child is learning in the classroom. So the way we approach this is like let's simplify this. Let's reduce some of the anxiety that parents face around teaching. Or helping be a coach to their child and making mistakes by one, ensuring that we provide data that tells them this is what your child is learning. And as they're learning, let's help them provide encouragement to their child. You know as a parent, you know, and you have young kids so you know this really well you tend to be nervous about your own child and their academic achievement because everybody's around you is telling you how their three-year-old can read and you know their four-year-old is now doing advanced calculus you know.

20:15.58

Dipesh

Yes, absolutely.

20:21.39

Sunil Gunderia

The reality is that every child especially at this age needs to be encouraged based on where they are and developmentally kids at this age are very different. And so one of the things we encourage the parent to do is celebrate your child's learning as it happens you know. One of the things that's going to do is really give the child a feeling of like wow not only does my teacher know what I'm learning but my parent does. And they care about it and it's important to them and when I learn something new they're there to celebrate it with me. And it's important to them. And then of course we provide opportunities and as examples of how a parent can provide support so that whatever learning that they did in the classroom and what they've learned the parent can say look - and you mentioned this example of counting to 20 - that. ‘hey, why don't you do that with me?’ And when you do that with me say, ‘Wow that's an amazing job’ or when you're out in the car or in the grocery store, ‘Hey, let's count to 20 together I know you've just learned this’. So what it helps is that it reinforces the parents' engagement and it also allows the child to see that the learning that they've done in their classroom is applicable to the world around them. And it helps them understand that world better and this is what you know creates a really solid ecosystem for learning where and makes the learning less about - this is something that happens in the classroom where I'm graded on it, and more about - this is important to my life and what's around me and everybody that cares about me cares about what I'm learning.

22:02.38

Dipesh

This is amazing. The first thing that I'm going to do after I go home, Sunil, today is to celebrate my child's learning. I think we obviously acknowledge it, but the way you said it. It. It just made so much sense, that each child is on its own journey and you have to take care of their journey, rather than trying to enforce something that's not a part of the journey. So that's amazing. I'll just ask one question before I move to a broader topic is - when you spoke about tools and teacher analytics, how do you make sure that teachers get the relevant data and and are not overwhelmed by a lot of data and dashboards? Because you know teachers see so many data points. So what is it? How do you do that?

22:56.70

Sunil Gunderia

In the same way, we would design for a student we've designed for teachers and one of the certificates we recently earned from Digital Promise which is a 501(c)(3) nonprofit here in the US that was created by the government to help promote better instructional practice is a teacher informed certificate that we've actually brought teachers into our development process. Not only for the student experience but for their own experience with our dashboards. So what I'm talking about is the same, the design research - you spend time with teachers and say, ‘Look here's our dashboards. How can we make this the most relevant to you so that you can not only understand where everyone in your classroom is, but also how you can use it to better differentiate your instruction?’ You bring your users into your development process and that is really quite key to ensuring that you're going to optimize usage of things like dashboards.

24:07.75

Dipesh

Great. So you know, thank you for providing this. I take a few steps back and I move to a question on AI. I think you work with a lot of institutes. So what I want to understand is a lot of institutes are definitely have the intent to adopt AI and are in the process of doing that. What are some of your ground experiences? What have you seen when it comes to institutions adopting AI or adaptive learning and what are some of the challenges that they face when they do that?

24:38.40

Sunil Gunderia

What has been encouraging is the speed at which districts and governments are looking at AI as a solution. And I think a lot of this has been driven by the needs of the workforce and what's going to be changing with AI. McKinsey recently put out a report that expects 30% of what we do at work to change by 2030. That it won't be needed anymore because AI can handle it and in some professions, it may be you know a lot more. And I do think that you know with interfaces like Claude, ChatGPT that we've surfaced a natural language interface to use AI. The power of AI as used by districts has been seen as not in terms of how can a teacher create lesson plans. How can they better communicate with the home environment? And there is a need to do this with consideration and you hear this from a lot of entities. One of the ones that I follow closely is the EdSafe AI Alliance which is looking at how do we instrument AI but consider the potential harms, including the biases that exist in a lot of these large language models because they rely on data that exist on the internet. And a lot of times that data is not equitable in terms of how it perceives people from different backgrounds and races and ethnic origins right? So we want to make sure that we are careful that we use this new technology in a meaningful way and for what it can power but also be considerate that it's not perfect, and that it's evolving. And it's something that we want to be very considerate of as we adopt it into our practices. We have seen I think 7 states so far in the US adopt early policies around edtech and I think we're going to see a lot more coming soon. We've also seen the Department of Education at the federal level actively look to put some guardrails, which are really important around AI into practice or consideration. So that we make the right decisions of how we use this new technology to really move the needle in terms of outcomes.

27:25.50

Dipesh

Great and you know you spoke about a very important topic about, you know safeguarding, putting guardrails. So I think that's great. I think one of the things with the emergence of AI is the ability to spot - an ability to identify - students who need additional support. Students who have learning difficulties. What is your take on that? How can AI enable that?

27:55.16

Sunil Gunderia

You know, similar to the system that I talked about - our Personalized Mastery Learning System - the opportunity with AI around that is, you know as you consider measurement, and there's a lot of research and theory behind how I believe an AI company or companies that are using AI in their edtech product should consider. So you know, similar to what we did. You know, you stand on top of the body of fantastic research that exists. And then you think about now I have a new technology. How might this technology help me and that help can be in a lot of different ways? Like how do I communicate to a teacher that a student may need additional help or that is struggling and you know, potentially a solution may have some sort of diagnostic that tells you that it could be something that is a disability that needs to be considered and look at you know, x, y and z in terms of how you handle it. Or it could be something like you know, being able to address - a teacher trying to communicate with a parent or caregiver at home that has a different heritage language or is more comfortable with a different heritage language. How they might use AI to take what a standard report is sent home about a disability to do one in the parents' language.

29:35.58

Dipesh

Great. I think that personally is very refreshing to see, you know, for people with different kinds of learning abilities to be able to be identified, and then given their own path to succeed. I think that's fantastic. Looking a little bit in the future, what trends or what possibilities are you most excited about when it comes to AI, adaptive learning, and the future of education?

30:06.27

Sunil Gunderia

I'm really excited about you know, one of the trends that we are definitely pushing to make into a movement is around evidence. And you know, as much as people are focusing on tools, our work as tech developers has to be about improving outcomes for the students that we serve. And of course, their teachers and their parents who are with us on this journey. And to me promoting that evidence and outcomes and ensuring that we're effective is really important. And I'm excited about some initiatives that are starting - one by the Southern Education Foundation around outcome-based contracting - and in a nutshell what that is, is that contracts are based where a developer like ourselves has some skin in the game in terms of a certain amount of our payment is contingent on actually achieving the outcomes that our products promise. To me that seems like it's a right way to line interest between developers like ourselves and the districts. That's one example of something I'm excited about. Another example that I'm excited about is that there's just a national movement around early childhood education. And you know, given this is an area we focused on and we think it's well beyond time and there's great evidence to show that investing in our kids from birth through when they're 8 years old has the largest return. Professor James Heckman at the University of Chicago has done some wonderful research that the return on investment is between 13 and 17% for every dollar spent, returns over the lifetime of the student, and with AI technology there's such an opportunity to ensure more of our students are ready for school and that when they get to school they can thrive and the parents understand what they can do in addition to what's happening at school to help the child along in their journey. The third part of this is I do really believe that AI and technology has a potential to do is really to move our systems and our institutional systems around education. What Todd Rose in his book ‘The End of Average’ described as a system that is standardized based on a time and seat to one that recognizes the needs of every student and to be able to personalize education as we've talked about and to ensure that we are providing to each student what they need to be successful. Not only academically, but in life.

33:08.10

Dipesh

Yeah, this, the book - ‘The End of Average’ is the second time in the last four days that somebody's recommended - that somebody's spoken about that book. I've already purchased it. I'm going to read that. I hear that it's amazing, especially when it comes to education and student journey. I'm definitely going to follow the outcome-based contracting because it does seem to be a step in the right direction. I'm sure there's a lot of work that's going to go to build this but it's definitely exciting to see the right measures being rewarded. So thank you for sharing that. Lastly, the last segment Sunil, we would like to focus on you as Chief Innovation Officer at the Age of Learning, somebody in the edtech space. A lot of tech companies today are trying to incorporate AI and new-age technology into their products. Based on your experience, what would be your recommendation to some of your peers and what kind of challenges can they encounter? How do they overcome that and kind of create products using technology as a backbone to deliver outcomes?

34:20.30

Sunil Gunderia

That's a great question and one that we think about all the time. You know AI has a lot of multiple purposes and you targeted really the most important. It is a tool as is any edtech product and that it's a tool for ensuring we help our systems, our school districts, and our parents create better outcomes for their children. And if you start there as your theory of change and you find your theory of change in terms of what do I want to do with my product that's really going to impact learning. And you know for us with My Math Academy and My Reading Academy, we already have 17 ESSA-aligned studies that showed that our product is effective at moving the needle and you know we're doing this in a way that works in the constraints of a classroom that is you know 15 to 20 minutes, you know about 3 to 4 times a week and we can get gains with that with our products doing that. I think as any edtech developer you want to think about - where is your product going to be used. What impact is it going to have? How are you going to measure that impact? And in a way that is, you know, valid and relevant and that you know a study basis and ESSA alignment is a way to do that. And I think that the other thing that is really interesting about the edtech tools is being able to use them now for understanding what research is out there. For a problem that you're trying to solve. So you know looking at there's tools now to say I can use AI to say - let's say I'm trying to solve the topic of student motivation all right? Well, what is the research basis for student motivation? What has been done at our great universities across the country already by professors who have really looked at this problem? And then alright - this is what the research says - well how might I incorporate that research into my product? How will that make it more effective? And then how could I test that in terms of evidence of that actual research being effective in the way I've deployed it in my product? And you continue to think about how do you use AI and AI technology to iterate on your product. To make it better. To make it more effective and ultimately to deliver hopefully what is a golden age of education because we have been able to shift our systems into something that delivers learning gains for all of our students not just a small percentage of them.

37:16.31

Dipesh

That's fantastic. So if I summarize you basically look at the research for the problem that you're trying to solve and use technology as a backbone to deliver impact. I think one of in one of the shows that I saw or I heard you said that you're in the business of creating impact. So I think that's where we think about impact and how what you do create an impact. This has been great. Thank you so much, Sunil. Before we wrap up, is there anything that I have not asked you and something like that you want to share?

37:46.11

Sunil Gunderia

You know you nailed it. We are in the business of impact, we're in the business of outcomes, and products like My Math Academy and My Reading Academy. We feel like we're creating something that is leading edge to drive those outcomes. That our students, our teachers, and our parents deserve. And frankly, what we're going to need to really prepare ourselves for the AI future ahead for all of us. So I'll just leave you with that.

38:18.11

Dipesh

Thank you so much, Sunil. Thank you for coming to our show. Where could folks find you if they have to connect with you?

38:23.20

Sunil Gunderia

Please feel free to reach out to us. You can find me at Age of Learning, I'm sunil@ageoflearning.com and you know please check out our website to learn more about our products and that's just ageoflearning.com.

38:40.11

Dipesh

All right! Thank you so much, Sunil. You really educated us about creating research-driven products that can deliver maximum impact using evidence. So thank you so much and we will stay in touch.

38:52.71

Sunil Gunderia

Thank you and really appreciate the opportunity to talk to you.

38:56.50

Dipesh

Thanks.