How AI Coaching Is Transforming STEM Classrooms with Smita Saxena & Alexia Lewis of Maestro

Smita Saxena is the Founder of Maestro, an AI coaching platform helping K-12 STEM teachers improve instruction and student outcomes. Alexia Lewis is an 8th Grade Algebra I teacher at KIPP Polaris and a Maestro coach whose students achieved remarkable gains using the platform.

💡 5 Things You'll Learn

1.  How AI coaching saves teachers hours every week. 
2.  Why human coaching + AI outperforms either alone. 
3.  How AI supports differentiated instruction. 
4.  Using AI to turn assessment data into action. 
5.  Practical ways AI can reduce teacher burnout. 

✨ Episode Highlights
[00:02:22] What Maestro's human-in-the-loop AI coaching model is.
[00:04:21] Saving 4–8 hours a week on lesson planning.
[00:08:20] Why AI works best alongside teacher coaches.
[00:21:28] Using AI to simplify data analysis and reteaching.
[00:23:38] How AI-powered instruction led to dramatic student growth.
[00:29:08] Scaling teacher coaching through experienced educators.
[00:35:41] The biggest surprises from implementing Maestro.

😎 Stay updated with Edtech Insiders! 

Follow us on our podcast, newsletter & LinkedIn here.

🎉 Presenting Sponsor/s:

Every year, K-12 districts and higher ed institutions spend over half a trillion dollars—but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts, and strategic plans, then helps you turn them into personalized outreach—fast. Win the deal before it hits the RFP stage. That’s how top edtech teams stay ahead.

Tuck Advisors was founded by entrepreneurs who built and sold their own companies, frustrated by other m and a firms, they created the one they wished they could have hired but couldn't find. One who understands what matters to founders and whose North Star KPI is the percentage of deals closed. If you're thinking of selling your ed tech company or buying one contact Tuck advisors now.

Innovation in preK to gray learning is powered by exceptional people. For over 15 years, EdTech companies of all sizes and stages have trusted HireEducation to find the talent that drives impact. When specific skills and experiences are mission-critical, HireEducation is a partner that delivers. Offering permanent, fractional, and executive recruitment, HireEducation knows the go-to-market talent you need. Learn more at HireEdu.com.

This season of Edtech Insiders is brought to you by Cooley LLP. Cooley is the go-to law firm for education and edtech innovators, offering industry-informed counsel across the 'pre-K to gray' spectrum. With a multidisciplinary approach and a powerful edtech ecosystem, Cooley helps shape the future of education.

Transcript

[00:00:00] Alex Sarlin: This season of EdTech Insiders is brought to you by Starbridge. Every year, K-12 districts and higher ed institutions spend over half a trillion dollars, but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts, and strategic plans, and then helps you turn them into personalized outreach fast.

Win the deal before it hits the RFP stage. That's how top edtech teams stay ahead. 

[00:00:31] Alexia Lewis: It's broad. It can be spread across any topic and be successful, be very successful. Just in math alone, seeing the data that my team has gotten back, seeing the growth that we've shown just working with Maestro, I mean, it's, it's crazy.

[00:00:52] Alex Sarlin: Welcome to EdTech Insiders, the top podcast covering the education technology industry. From funding rounds to impact to AI developments across early childhood, K-12, higher ed and work, you'll find it all here at EdTech Insiders. 

[00:01:08] Ben Kornell: Remember to subscribe to the pod, check out our newsletter and also our event calendar.

And to go deeper, check out EdTech Insiders+ where you can get premium content, access to our WhatsApp channel, early access to events and back channel insights from Alex and Ben. Hope you enjoy today's pod.

[00:01:32] Alex Sarlin: On EdTech Insiders today, we are talking to a fantastic pair of people, an entrepreneur and an educator that are working to improve mastery of students. It's incredibly important work. We're speaking to Smita Saxena. She's a founder of Maestro, which is an AI and coaching platform for K-12 STEM teachers, and she's seeing year one classrooms beating national growth norms by 200% to 400%.

Incredible. And joining her today is Alexia Lewis, an eighth grade algebra one teacher at KIPP Polaris in Houston, whose students nearly doubled their mastery scores in her first year teaching using Maestro. Really exciting to see you both. Smita Saxena and Alexia Lewis, welcome to EdTech Insiders. 

[00:02:18] Smita Saxena: Great to be here.

Thanks, Alex.

[00:02:20] Alexia Lewis: Thank you. I'm excited. 

[00:02:22] Alex Sarlin: I'm very excited to speak with both of you. So, so, um, Smita, let me start with you. For people who don't yet know what Maestro is and what you're offering to schools and to first-year teachers or teachers who are looking to upskill and improve mastery in their classroom, what are you doing with Maestro, and what makes it different than what's been on the market so far?

[00:02:40] Smita Saxena: Yeah. So like in a nutshell, Maestro is human-in-the-loop AI coaching. And so what that basically means is, as a teacher, you are still paired with a human coach, and we go through and we try to understand what your workflows are, which could be anything from lesson planning to your data analysis, to your reteaches, to any sort of communication that you're also doing internally, trying to see how we can make those more efficient using AI.

So for teachers that are part of our program today, the majority of their time that we see is going towards lesson planning. And so the first thing that we try to come in with is understand how they go about their lesson planning process, understand what sources they're pulling from, where the gaps are in their curriculum.

Keep in mind, we only work with Title I charter schools, so 40% usually of the students coming into these classrooms are testing three or more grade levels below. So for a lot of these teachers, the challenge really is closing the gap for where their students are today, but then also closing the gap between the curriculum and the assessments.

And this is what AI is incredible at, is creating those supplemental materials that those students can then go through and use. And we'll talk through more examples of how teachers are using it as we kind of get into it as well. But this is kind of- Yeah ... what we 

[00:03:56] Alex Sarlin: do. And we have the privilege of having an educator in the to, to speak to it right now.

So let me ask you, Alexia, from your perspective, when you were training with Maestro and using Maestro in your classroom, tell us a little bit about what that experience looked like. What did it-- How did it support lesson planning? How did it create supplementals, and what does it do to sort of enhance your practice such that you can see incredible growth in even your first year of teaching?

[00:04:21] Alexia Lewis: Okay, so this is my first year teaching with Maestro, so I'm actually going into my fifth year of teaching, and I can 110% say that it has taken So it's given me back my time. I went from spending Four to eight hours a week just on lesson plans outside of the classroom. And with Maestro, they've taken that completely away because now they're internalizing my lesson plans, and we're able to collab, and I'm able to tell them if I want work for small group or intervention.

I'm also getting these tier one, tier two, tier three assignments, so I'm able to start with the basics for my students and see, okay, are we able to get past this tier one assignment? And as they get past the assignment, you know, of course, it gets progressively harder. And I'm able to get these mastery levels by making sure that all the X's are crossed off.

Even in the assignments that Maestro has created, that we've been able to go back and edit to make sure that it helps my students. I'm able to put what we have, what we call a I do, you do, we do. Well, I do, we do, you do, where now they see my work, then they're able to practice my work by filling in the blanks, and then they have their own independent practice.

And because we work with a group of kids now where I will say COVID affected our education system a lot, so we're able to now play catch up and fill in those gaps. And we know algebra is the foundation of maths, so now they're able to understand this, so they can take it and run with it in other courses.

[00:05:54] Alex Sarlin: Absolutely. A- algebra is this gateway course to, to al- so much of, of math, and it's also famous for being a, a blocker course, a course that a lot of students really struggle with. So that differentiation, that multi-tier support is, is vital for catching people up. And you mentioned the, the pandemic learning loss that we've seen in all the NAEP scores.

We've seen, you know, national learning loss for, for years now. It's only starting to reverse now i- in math. It's powerful. I wanna get into the human in the loop side of this as well, because it's I-- obviously a really key part. But you hear Alexia mentioning four to eight hours of lesson planning that was returned to her, and you've, you've named that, you know, one of the most important measures of AI return on investment is basically how much time is returned to educators that they can use for things that they lean in on, on, on, uh, actually supporting students rather than doing behind-the-scenes work or planning or administrative work.

I'd love to hear, Smita, you talk about that writ large. And then, Alexia, you've already begun to talk about it, but I'd love to hear you talk about it as well, what it looked like in the classroom. 

[00:06:53] Smita Saxena: So I'll rewind a little bit 'cause kind of how I fell into this. So I live in San Francisco, and I was a math major in, in undergrad and then through, through grad school as well.

And when I stepped away from my last company, I started volunteering at a, at a public school in San Francisco, elementary school, math. And it was the first time I kind of got a view into what a teacher's day looks like. And it kind of blew my mind that within the first thirty minutes, you're touching basically thirty different tools.

And coming from, like, our world of tech, where everything is, like, hyper-efficient, and there's, like, this whole idea of, like, an employee experience, right? You're set up to succeed because- There's dollars on the line here. And, and also in terms of like the systems that you kind of have in place, they work for you versus like against you.

So my kind of insight into this, and I spent, you know, a bunch of time with the teachers over here as well, was you spend a lot of time doing, you know, duplicative work and work that isn't also tying back to what is impacting your students because that feedback loop doesn't exist. And, you know, we live and die off of feedback loops in, in tech and product world.

And what I also kind of saw on the other side of this is the work that teachers were doing manually was 'cause they had to do it manually 'cause the tools did not exist in terms of AI just was not at the place where we could go in and generate the materials at the scale that we needed to. So I always thought of it from the perspective of like, how do we reallocate your time from this like repetitive work that you're doing towards actually being more thoughtful about how you're teaching your students?

Like thinking about like the why behind what is not working, and that takes time. It takes space. You have to like really kind of be able to take a step back from your day-to-day in order to have those insights. And so that's why I think the human and label peach- piece is so important. Where you have... You can have-- You have so many tools that can generate materials for you.

There's no shortage of that. The question though of like figuring out what you need, why you need it, and how you're gonna implement it, that's where the human coaching piece comes in And so when we first started the program, it was actually me as the coach, which was interesting 'cause I don't have, like, an education background, so I was coming at it from, like, the tech and product piece.

So like, "Let's poke holes here. Let's look at the data. Like, why do you think this is working or this isn't working?" And then about six to eight weeks into the program, we had enough data to be like, "Wow, we actually have teachers that are outperforming, Alexia being one of them. And can we actually take some of the six to, six to eight hours a week that we're saving you, can we turn you into a coach where you can go through and start working with some of the other teachers that we have?"

And Alexia is actually one of our, our star coaches. So we can talk about that as well. But it's you need both pieces. You can't just have the materials because it's the what. The how, why, when, you know, is all kind of the human piece of it. 

[00:09:45] Alex Sarlin: Absolutely. And so, so Alexia, from your perspective, you had four to eight hours of lesson planning a week, you know, and studies show so many of teachers' hours are spent on administrative tasks or writing to parents or, you know, there's just so many different things that fall onto their plate that are non-e-educational.

They're not instructional minutes necessarily. You get all this time back. What are you doing with the time to enhance the mastery of your students as well as to coach and pass it forward, pay it forward and, and, and help other educators support their students? 

[00:10:15] Alexia Lewis: You know, now that I got a little time back, I'm able to do some fun things outside of work, right?

So with the coaching piece, so what I like to do is collaborative thinking. So anytime I step into a scenario with one of the teachers that I'm coaching, I like to make sure that they feel heard and that we can put our brains together. So how these coaching meetings sometimes go is me telling them what I do in the classroom, how we got the data to go from here to here.

What are you doing in the classroom? Now let's put these think pieces together and see how we can make this classroom better as a whole, whether it be from small group, working with, you know, your special education kids, working with your regular population kids, or for example, at my job where everybody's in one setting.

So what are we doing to make sure that everybody's being heard, everybody is being-- is able to adapt, and everybody is learning? Because you can't move on if only 20% of your class is understanding what you're learning and the other 80% is kind of depending on that 20%, which I've seen a lot. So what we do is, when I'm in these meetings with these teachers, it's making sure that we're reaching that 80% that's not talking.

We're making sure that everybody feels heard and that the classroom is a comfortable environment because we know nowadays, shoot, even when we were kids, kids were mean. So you know, if you were too smart, if you were too smart, you, you know, you tried not to say much, but now we're setting up these classroom environments where everybody's comfortable to be heard.

Let's make it fun. What can we do to where the classroom isn't boring? Students should look forward to coming to your class. So what are we going to do to make sure that this is happening? What are we gonna do to make sure that we're catering to everybody's needs and, and what are we going to do to make sure that all of these kids are STAAR ready?

[00:12:01] Smita Saxena: I think the one thing I'll add, and Alexia, you've done such a great job of this, is figuring out, like, you know, for some of the teachers you were coaching the types of projects that they were looking for in order to kind of keep the students engaged versus for certain types of teachers it was like, "Hey, we wanna do more gallery walks or certain types of scaffolding worksheets to get them to where they needed to."

So I think those pieces, just the, the different modalities that I think the coaches kind of brought to us was what really helped us, you know, personalize this on many levels. Yeah. 

[00:12:34] Alexia Lewis: Just to, um, backdoor off of what she just said, we also made sure to create a minute-by-minute schedule. So like, there, that way the teacher's not feeling like they're up and having to teach the whole time and that they're not on their feet because, you know, uh, we hear a lot of teachers complain about the fact that they're up the whole time, like they're burnt out before Christmas even comes.

So another thing we try to avoid is that burnout feeling that you feel right before midterms, right before STAAR, right before finals. So we do oftentimes create a minute-by-minute schedule as well, just so that you know. With this material, if you implement it correctly and if you implement it to make sure that everybody's learning, you should only be standing up for 15 and no longer than 20 minutes 

[00:13:20] Alex Sarlin: I hear this really interesting theme through both of your answers about sort of adapting or personalizing or sort of adjusting in, in real time the instruction, the, the avoiding burnout at really high-stress periods of the year, like right before, you know, major assessments like the STAAR or, or early in the year.

Adjusting student materials to their tier, or whether the, they're the 20% of students who didn't understand the, the core math concept, even though the class is moving forward, you need materials to support remediation. And I feel like this is a core thing that AI can provide, is that sort of it can, it can do many versions of things, including ones at different levels.

But what I'm hearing from you, Alexia, is that one of the things that the human coaches provide is, you know, engagement, is sympathy and empathy with the teachers about what they're actually dealing with, is, you know, being able to sit down and think really critically about this particular classroom, this particular set of kids, what the teacher philosophy is, and sort of bring all of this very human context.

And I'd love to hear you both talk about sort of how the, the, the strengths of the AI, the ability to generate, you know, infinite amounts of content and to tailor it and transform it and all of that, work really alongside the human capabilities of trained educators who say, "Oh, yeah, I, I know where you've been.

I want you to be stand-- You know, I'm gonna make you a lesson plan where you're sitting down for half of the period." Like, that's a really human feature. I'd love to hear you both talk about how the sort of the two pieces come together to create a, an experience that's greater than the sum of its parts.

It's ideally better than, uh, just having a human coach and better than having a just, an AI professional development system. Uh, Smita, let me start with you. 

[00:14:55] Smita Saxena: Oh my gosh, yes. So initially, you know, given my background and, um, you know, where I thought this would just be a tech-only tool, where it would just be an app that you would go to and you could go through and like generate materials, right?

And then as I started, you know, working with the teachers, it became very obvious to me very quickly that that human angle is absolutely necessary. And I call it like the centaur mindset. It actually comes from the world of chess where, I'm not sure if you're familiar with this, but when Big Blue, you know, beat the first grandmaster, there was this whole set of chess games that were happening where it was human plus machine versus human plus machine.

And they realized that those were-- a human plus machine would almost always beat a machine only or a human only. And that kind of applies across many different areas, this being certainly one of those, those places. But what I think was very, very impressive, you know, with the teachers that we worked with was like, and the coaches, is how quickly they embraced it and how much creativity was being unleashed in terms of like what you could do.

And also a sense of relief in terms of I finally have resources to meet my students where they are And I don't have to kill myself over it, right? Like, I can actually, like, do my job, enjoy it, and see results, was like... I think that was when we got, like, our aha moment of, oh, yeah, this could work. And seeing that there are coaches that are within the system, you know, we're, we're, you know, we're obviously kind of adding income over there as well for them, which we, which we love.

And so it's like we wanna keep the best teachers in the classroom. We wanna make their jobs easier, and we wanna bring everyone up with them, right? And this felt like one of those unique ways that we could do it. I'm sure there's several ways to do it, but the fact that we were able to hit all of those goals as quickly as we did was quite surprising and, like, you know, honestly, like, very, very rewarding.

But I'll let Alexia talk about it 'cause she's actually in the classroom. So, yeah. 

[00:16:49] Alexia Lewis: So do you mind restating your question one more time? 

[00:16:52] Alex Sarlin: Sure. I, I'm just saying that, you know, in the past you might have had a human coach working with a teacher and you'd say, "Oh, I understand these kids. You just need to have like, uh, four different versions of this lesson for different kids."

And the teacher'd go, "Great, but I don't have time to do four different versions of this lesson," right? And the AI can handle that. If you just have AI supporting a teacher, it might say, "I can make four different lessons, but I don't actually understand your context in your classroom. I can't empathize with you with, as a teacher.

I can't ma- I can't sit and plan with you." So it feels like the human plus the AI is better than either one alone, and I'd love to hear what that looks like for you in the classroom. 

[00:17:27] Alexia Lewis: So as someone who just had a coach, you know, it's easier said than done. Let's put it like that. So you can tell me that I need to do this, this, this, this and this, but there's only 24 hours in a day, and, you know, majority of my day is spent here.

So with this human plus AI kinda interaction and how I do it with, uh, my team and how the assignments are created for me, okay, so let's start with step one. What do you think these kids are gonna need to know most? Coming from my class, they need to know how to work it on a calculator. So I am putting how to do it on a calculator on this worksheet.

Now, in the next section, I'm telling you how we're gonna plug it into the calculator, what your answer should look like. Then on step three, I'm telling you how to write the answer. So now it's me on the board for about 15 minutes showing them an example of, "Okay, so guys, let's look at section one. What is it telling us to do with the calculator?

What are we looking for?" They're able to tell me that without them having any knowledge of what we're about to do, solely because the paper's providing it, because of what AI is creating is providing it. So if they have what they need directly in front of them, it sticks rather than me trying to get them to memorize it.

I mean, before AI, I was coming up with songs, like trying to do whatever I could to make sure that the kids understood it. You know, these formulas that aren't provided on these formula sheets or showing them how to graph, which is something that once you learn it, you have to remember it or if you forget it, you know, you're cooked.

So like it, this worksheet now that I, I have, I've even had parents tell me, "Ms. Lewis, I ... " When they brought this worksheet home and I looked over it myself, "Shoot, you taught me." So I was able to help them some more because it re- it reminded me of what this was. So it's helping everybody in like the scenario.

It's helping all the teachers. It's taken away so much talk, and now that we can work, because you can see it on the paper, so now you just need to implement these steps 'cause it's, I'm give- making sure that I set you up where it's step one, step two, step three, step four What it'll look like on the graph.

So I'm set, I've set you up. Now, now I'm able to go back and see, okay, who doesn't understand? Who still does not understand even with the instructions in front of you? Then you realize you have such a small group, just a small, small group. It's always gonna be, you know, somebody who just really doesn't understand, and now I'm able to kind of sit down with them.

And then I do a lot of peer teaching in my class, and I tell my team, the people I coach, to also do peer teaching 'cause students may understand it better from a classmate than they will from us. Sometimes they close their ears when an adult tries to come and correct them, so they may do better learning from a classmate, learning from their friend.

And now it's become cool. Now it's become fun. Now we can talk as long as I know that what we're talking about has something to do with what we're learning. 

[00:20:23] Alex Sarlin: Yeah. It's so interesting to hear, you know, what I'm sort of hearing you say is that as a teacher, you're getting all of these supports, right? I mean, p- peer teaching is a support because you can have peer, you, you-- the students can sort of help, uh, remediate each other, help support and, and, and sympathize, and they just learned it, and then they can pass it on in their own words.

But you're also obviously getting so much support from the AI-generated materials, from the human coach who's, who's sort of helped you set up. It's like teaching can be a very solitary profession, and the way you described it, it doesn't sound solitary at all. It feels like you have sort of a whole world of resources surrounding you to make sure that things work.

You have the data, you have, you know- It's just it, it feels a lot-- And when you mentioned earlier that it, it can help reduce burnout or help support in those really, really stressful periods, I think that, that you can sort of see why it would if you're, if you're coming into a classroom feeling like you have all of these materials, you have, uh, students who can support, you have all...

It's a much more exciting, I think, vision of a teacher day than what we often envision, and, uh, it seems like Maestro is really leading 

[00:21:26] Alexia Lewis: to that. Yes, definitely. 

[00:21:28] Smita Saxena: I'm glad you brought up the data piece too, because that's the feedback loop part of this, and something that we're, again, you know, when I first started getting into this space, was surprised by like how much of the work of analyzing the data goes towards teachers, right?

Where I'm like, I don't-- I- it's... You know, I guess for me as an outsider, it just felt a little bizarre that you already have such a busy day with all your students, and you're doing lesson planning, and now suddenly you need to kind of context switch and also understand your data with very limited amount of time, and then figure out what to do with the data.

And so to me, like, and I think these are things, again, coming from my world where you have teams that do that, right? Like if you're in the product world, you have a data team that's gonna give you your readout, you make decisions, and you go, you move right on it. So having to do all that work, just not enough time in the day to do it well.

And having to expect teachers to do that on almost like a daily basis felt-- it just didn't feel doable. And so as part of our program, as part of the coaching that we do, is we do look at that data on your behalf. We, you know, create the misconception list of looking at your data and looking at what the performance looks like.

And then we can do, we can just literally give you your reteaches, being like, "This is what we see." And as a coach, you can kind of pressure test this being like, "Are you, you know, are you seeing the same thing that AI is or are there things that you think we're missing here?" But it kind of takes what used to be hours and like bringing it down to like fifteen minutes.

So you're spending so much more time just figuring out what you want in your reteach rather than trying to struggle with the data on like, what does this actually mean? And again, like I think we're very lucky that we have these models that are as powerful as they are because, I mean, this data, some of it is not in the best shape, right?

You're just like, I mean, coming again from our world where everything is normalized and it's, and it's, and it's, you know, it's designed for you to actually succeed. There's some of this stuff where I'm like, yo, you're going through like 10 different screens to just figure out like how your students are doing.

This is like inefficient. And so the fact that we can just kind of do that in 15 minutes, I think was one of the, the cornerstones to how we thought about our product strategy was like, this has to go down to 15 minutes. We need to spend more time on reteaches. 

[00:23:36] Alex Sarlin: 100%. Oh, go ahead, Alexia. 

[00:23:38] Alexia Lewis: Oh, yeah, and this, just to kind of piggyback off what she said.

So with the data, um, you know, something I always tell my team is that teachers make money. You make more money with higher data. The higher your data, the more money you can make. People think, you know, teaching is just, you stick at one price point. But we are a world of stipends, right? A world of bonuses.

So for example, this year, before I started working with, with Maestro, my students take this test called IA, which is an interim assessment. They have a one and they have a two that they take. So one that they take in the fall, the last one they take in spring, and it's just kind of measuring their growth.

Before I started working with Maestro, my students were testing at a second-grade level, third-grade level, fourth-grade level. Yeah, these are eighth-grade students. My lowest student tested at like a kindergarten. So to say that I, I started so low, I jumped up almost 400%, which I genuinely didn't even know was possible.

Almost 400% by implementing a review sheet that Maestro, that AI generated for me based on my previous data. So we're able to Get this data, internalize it, let AI do what it does, and give me what I need to make sure that my kids are doing what they need to do on these tests. 

[00:24:59] Alex Sarlin: That's incredible. I wanna a-ask a follow-up question on that, and it's just sort of doubling down on the same idea.

You know, Smita, you're mentioning that, you know, the teacher-- we all know, teachers sort of catch all of the different work that, you know, they have to do, the lesson planning, the lesson delivery, the remediation, the extensions, translation when there's, when there's students who don't speak, you know, speak different languages in the classroom.

You have parent communication, uh, data analysis, uh, obviously a big one that we're talking about here. Sometimes there's training for that, sometimes there isn't, but there certainly is not enough time for all of that, and that's just sort of like a given. And it feels like, you know, one of the things I'm hearing you both say is that AI can reach across all of these different tools.

You mentioned the 30 different tools, uh, Smita, and, and, and Alexia, you sort of mentioned all these different types of assessments, all of these different tools that, uh, you could b- that you're using in the classroom, and sort of piece-- put the pieces together and then turn it into a usable format.

Actionable insights, small groups to, to remediate, peers to teach one another, or, or worksheets or review guides, like you just said, review sheets or calculator instructions. And it feels like, you know, this is exactly the dream I think we've all had in the edtech world about what AI could do, is it can take all of the complex work that there's not enough time to do and make it simple and easy and spit out something that is incredibly valuable for a teacher that makes the teacher's day better, it makes the teacher's-- makes the results obviously better.

It makes the teacher make more money, to your point, Alexia, right? If they're, if they're seeing really amazing results. It's like teachers are, have been so overwhelmed, it's one of the highest burnout professions, and I'm just hearing all of these exciting stories about how AI can, can lift the burden. So Alexia, I guess my question for you is, do you see Maestro or AI tools in general being able to sort of take the data from multiple systems or take the signals from, from you're, you're getting sort of incoming parent notes and incoming IEPs and all these different things, and sort of m-synthesize and make it easier for you to turn, to, to actually turn it into an instructional plan rather than just swimming in incomplete, to your point, Smita, data?

[00:27:04] Alexia Lewis: Yes. Oh, 100%. So for example, a lot of students on their IEPs have guided notes right? Now with AI, I'm able to create these guided notes on these lessons, but I'm also realizing that it doesn't need to be just for my IEP kids or, you know, my 504 kids. This is helping everybody. So you know what? Let's go ahead and just give it to everybody.

So now everybody has these guided notes. They have a step-by-step plan. They have a fill-in-the-blanks version. The only difference is that some students may get extra time than others, but besides that, with- All of this extra help and all of this extra product on the work, everybody's understanding. Like, you would think that people would struggle with quadratic equations, which is something that I've noticed kids have struggled with for years, for example, and then they get this work that AI, that Maestro has created for us, and I mean, quadratic equations turned out to be one of my highest data tests.

P- you know, so it's like it, it's working. We're gonna stick with what works, and if everybody needs these guided notes assignments that I just made for my IEP kids, you know what? Everybody gets it. 

[00:28:14] Alex Sarlin: Incredible. We did a session with some special ed providers a few months ago, and, and one of the big themes was exactly what you just said.

E- everybody should have individualized education, right? The things that work for students who have IEPs, a- and we're in a world where suddenly everybody can have individualized instruction. They can all get those supports. I love hearing you say that. That's so exciting. So Smita, let me pass this back to you because, you know, as you grow Maestro, you're, you're relatively young as a company.

I, you know, I think you're seeing really outsized results, including y- you know, a hund- 200, 400% growth in mastery for, for, in the first year of usage. How are you planning to... Y- you know, s- it's easy to scale the technology. That technology scales. But how do you scale the human coaching? What is the plan to make sure that you have high quality, really, you know, expert coaches that can be the human in the loop as you get bigger and bigger and reach more and more districts?

[00:29:08] Smita Saxena: Yeah. So I think what we learned kind of from like our pilot here is that You know, we kind of have to-- it's always teachers first in our approach. And so the way we're kind of thinking about our scaling is, you know, we're gonna obviously have some coaches that are on the team who are gonna be going into new districts, working with teachers the same way I did with the K- with KIPP and Spring and a couple other districts in, in Houston.

But the goal is ultimately to graduate teachers to becoming coaches within our program. And coming up with like a flexible model where you can coach up to like, you know, five teachers. So if you have time to just coach one or two, go for that, but you know, you can go up to five. And also kind of-- and one of the things that we definitely wanna try out over the next year is like group sessions.

So putting a number of teachers in one, one group and like, you know, on a maybe once a month or twice a month where they can discuss strategies, especially if they're kind of at the same level or, you know, dealing with similar problems, which keep in mind, all of our coaching sessions are recorded. And so we get-- we run all of this through our models as well to basically understand what are the traits of like what a good coach looks like, when a teacher is ready to be graduated to being a coach, what kind of teachers are facing some of the same issues, and what are the trends that we're able to get.

And a lot of that kind of actually gives us a sense of like predictions around, hey, this is how mature a certain cohort is. We can like move 'em on to this piece. We see ourselves more as like setting up systems here where we wanna keep the best teachers in the classroom. And how quickly and like what do we need to put in place in order to make that happen is what a lot of our research is gonna be going into over the next like few years.

But from the early signs, we're also very happy to see that our coaches continue to excel within our cohorts. So our coaches did not see a drop in performance as they took on teachers. And that was like a worry that we had was like, hey, if you're taking time away, is that gonna impact your, your thing?

And so we were watching that very closely as well, and we saw that all of our coaches still outperform their cohorts. All these learnings I think are gonna go into like how we scale that human piece 

[00:31:18] Alex Sarlin: Yeah. I love the emphasis on, on community, on creating groups and, and, and sort of paying it forward, the idea that, that teachers who are succeeding and seeing amazing mastery growth in their classroom can become coaches.

And then, of course, as you say, you can use-- even use the AI to identify the, the behaviors and, and, uh, you know, what, what, what coaches do. That's really exciting to see. I can imagine a world, you know, Alexia, where you have, you have a set of coaches who you've trained, and they have a set of five coaches who they've trained, and they have a set of five, and it becomes this, this great, you know, pyramid of expertise, which is really, really, really exciting.

Last question here. We're coming to the end, but I think we have a really-- this is such an interesting model. I noticed that you're, you're starting with STEM and, you know, Alexia, you are a algebra teacher. You're a math teacher. Th- math is a hugely important topic for this. I'm curious if you feel like this model could work in other subjects.

Have you had people come over from the English department? Have they-- people from the science department said, "Oh, I want a coach," or, "I wanna be a coach," or, "I really want this"? 

[00:32:16] Alexia Lewis: Yes, 100%. So it's crazy because our whole English team is like, "Why are we left out the loop? It's science and math and everybody's focusing on STEM right now."

But no, I, I could see it helping literally every cohort from elementary, middle, and high school because it's applicable. So you're able to understand it and dissect it, and it doesn't matter what subject or topic it's in. I feel like it would especially excel in history because it would make memorization kind of looked at in, in a different aspect in a sense.

And I've taught, I taught science my first two years of teaching, and I've seen how it turns science into something fun. So you're still learning, you're still able to understand because, you know, science is application based as well the older kids get. It's not about memorization. It's not chemistry where you have to memorize it, 'cause even chemistry is still applica- application.

But it's broad. It can be spread across any topic and be successful. Be very successful. Just in math alone, seeing the data that my team has gotten back, seeing the growth that we've shown just working with Maestro, I mean, it's, it's crazy. So, you know, yeah, the RLA team at our campus is a little jealous, but but no, I got them one year.

Give us, give, give us some time. 

[00:33:38] Alex Sarlin: Smita, where are you, when, when are you heading to reading and language 

[00:33:41] Smita Saxena: arts and- I don't know when we'll be doing all of those things. So it's, it's so interesting. So we picked math and science partially because, I mean, I am a math major, right? So I think I understood, I had that felt sense of what good math looks like or good math instruction looks like.

Had some opinions around that. But also you have to kind of keep in mind the constraints of their models today is hallucinations, right? This is something that it continues to be a problem, will continue to be a problem, and is becoming more insidious as the models get more and more comp... One of the things that we do have with math and science is objective truth, right?

So we can go through and verify that these worksheets are actually accurate. We still have, you know, some issues that may slip here and there, and we, we actually hear that students catch them, which makes us happy that they've learned something. But at the same time, like that is something, you know, part of the problems with generalized AI models is like you just go off and generate a worksheet.

You don't actually know if it's accurate. And we've done a lot of work on, in our stuff internally to be like everything we generated has gone through a bunch of different checks, which we can do in math and science. I think we need to come up with a similar framework in other spaces, so whether it's ELA or history, et cetera, where we can make sure we are subscribing to the same level of quality.

And again, it's again like an area of research for us, and when we feel like we have that confidence, we will certainly be launching in those spaces. Yes 

[00:34:59] Alex Sarlin: That'll be exciting to see. Fascinating conversation. Let me just give you each a, a moment for just a last word. You know, as somebody who's sort of listening to this discussion and seeing, oh, this is such a, a humanizing and sort of really exciting way to see how AI can improve outcomes in a classroom.

It looks very different than some of the other edtech tools out there. It's not a AI tutor that goes directly to students and talks to them. It's not character that you talk to. It's like it's, it's really working very, very directly with the teacher and with the teacher coach. Let me give each of you just a moment to sum up, you know, what is the most surprising thing that you've experienced in your time working with Maestro?

Uh, Smita, let me start with you, and Alexia, I'll give you the last word. 

[00:35:41] Smita Saxena: I think the most surprising thing for me, and I say this every single time I tell this story, is like how much and how well the teachers embrace the program. To be honest, like I am an outsider, right? And I wasn't sure if teachers would just be like, "Hey, this is just another tech person trying to come in and like solve our problems for us," right?

And I was hypersensitive to that coming into this and wanting to make sure that we were building something very inclusive and we're, again, I keep saying teachers first because truly, I, I think the only way this works is if you have happy teachers. Happy teachers, happy students. And so that kind of took me by surprise.

And you know, we started off with just five or six people, teachers, you know, like six months ago, and now we're over 50 in one area, and we are-- you know, we don't have any salespeople. This has just grown from like word of mouth, and the fact that teachers are kind of recruiting other teachers and hearing about this and wanting to join the program has just been very, very validating, right?

But also huge surprise, you know, 'cause it's still AI and, you know, comes with its own set of like cloud or shadow, as you put it. And so yeah, it's been a very pleasant surprise. 

[00:36:47] Alex Sarlin: That's great. And Alexia, how about you? What's been the biggest surprise? 

[00:36:52] Alexia Lewis: So I'm gonna go with two. So first, how my students adapted to it almost immediately because, you know, I did bring it in mid-year, so from August to December we were learning based on just strictly textbooks and outside information that I was using, and me standing up almost 45 minutes out of a 85-minute class period to just that complete switch up, and my students adapted to it almost immediately.

They loved it. So them being the teachable kids that they are, it made it very easy, like the transition. And then the second one, the second surprise would be how fast data increased. Because, you know, you hear about the stuff, and I know that like we were the first group of teachers before I even became a coach, we were the first group of teachers to work under Maestro, and seeing that almost immediate jump, like the moment we started working with it, my data jumped tremendously.

I never backtracked ever again. So, you know, I'm scared it set a, a higher standard now, so like they'll expect it from this, you know, next group of kids, and I know we'll do it. But just the immediate jump, the congratulatory emails that I was receiving from district, from outside campuses, from outside administration who are simply just sitting there watching my data, trying to figure out what we did.

And I'm like, "Okay, I got the secret sauce." You know? So it was crazy. It was a huge surprise how fast my students adapted and how fast data grew. I mean, it was amazing. I'm very grateful for the program. I think it couldn't have came at a better time, and I, I love it. 

[00:38:29] Alex Sarlin: Amazing. And, and making quadratic equations your like one of the top subjects, and I don't know if I've ever-- The quadratic formula, quadratic equations, I've never heard that.

It, it's like that is like such a famously problematic subject for, for teaching. It's- 

[00:38:43] Alexia Lewis: It is. 

[00:38:44] Alex Sarlin: Amazing. This has been such a fun conversation. Um, Smita Saxena is founder of Maestro, which is AI and coaching for K-12 STEM teachers for now, and Alexia Lewis is an eighth grade Algebra 1 teacher at KIPP Polaris in Houston, uh, using Maestro, and she's seen doubles or up to 400% in some subjects, a mastery improvement in Maestro using her-- in her first year using Maestro.

Thank you both so much for being here with us on EdTech Insiders. 

[00:39:12] Smita Saxena: Thank you.

[00:39:13] Alexia Lewis: Thank you. 

[00:39:14] Alex Sarlin: Thanks for listening to this episode of EdTech Insiders. 

If you like 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.

This season of EdTech Insiders is brought to you by Cooley LLP. Cooley is the go-to law firm for education and edtech innovators, offering industry-informed counsel across the pre-K to gray spectrum. With a multidisciplinary approach and a powerful edtech ecosystem, Cooley helps shape the future of education