Make It Mindful: Insights for Global Learning

Seth Fleischauer (00:00.885)
Hello everyone. And welcome to make it mindful. The podcast exploring how students learn who they're becoming and how global experiences, human development and emerging technologies are reshaping teaching. I'm Seth Fleishauer, former classroom teacher and founder of an international learning company, focusing on global learning. Each episode features deep conversations with educators, researchers, psychologists, and school leaders who are rethinking teaching and learning. Together, we examine how students make sense of a rapidly changing world and how technology can support.

Rather than replace the heart of human edu see, did it. I almost got through, Lucas. I'm going to try that again. Together. We examine how students make sense of a rapidly changing world and how technology can support rather than replace the human heart of education. And that is right in line with the work of today's guest, Stephanie for now, the founder and CEO of school ops AI and a seasoned educator whose career spans teaching, instructional coaching, and school leadership.

A Pahara Institute fellow and former Teach for America Corps member, Stephanie holds degrees from Georgetown and Stanford and spent her career working to improve educational systems and student outcomes. Before launching SchoolOps AI, Stephanie founded Fair Schools, where she partnered with principals to strengthen school environments through research-backed practices, social-emotional learning, and meaningful family engagement. She also led literacy and system-level initiatives at Shusterman Family Philanthropies,

focusing on improving opportunities for black and Latin X students and previously served as principal with Rocketship public schools. The founder of which Preston Smith was on this podcast about that 40 episodes ago. Now at school ops AI Stephanie is bringing her deep expertise in school leadership into the emerging world of artificial intelligence. Her platform streamlines school operations by integrating quantitative and qualitative data into actionable insights.

giving principals more time to invest in what matters most students, families, and staff. In this episode, we explore the future of school leadership in an AI augmented world, what principals actually need, how smarter systems can reduce operational burden and how technology can help create healthier, more human centered schools. Stephanie, welcome to the podcast.

Stephanie Frenel (02:15.416)
Thank you. Thanks for having me.

Seth Fleischauer (02:17.497)
I am so excited for this conversation. We met at a world savvy event in Oakland last month. world savvy other guests that have been on this podcast, KK Nieman and, Harbin Porter were on the podcast. I don't know, maybe 30 episodes ago. I should probably have these numbers ready for people. but, that's also a good episode to listen to and.

this event that we were at was about AI and education. sat next to each other. You started talking about your, your product here, school ops AI. And I was absolutely fascinated because I've spent so much time in this world of skills that are very difficult to teach and difficult to assess. And oftentimes are best assessed with qualitative data. You can get the quantitative data, but like the real meat of it, the real story is in the qualitative data. Yet qualitative data is very difficult.

to leverage, tell stories. People want numbers, right? And so I'm wondering, what led you to starting SchoolOps AI? Was it that tension, that same tension that you were experiencing of there's all this good stuff that's being taught out there that can be assessed qualitatively, or was there some other story that led you here?

Stephanie Frenel (03:29.164)
Yeah, and I'll just start a little bit with like the context. So, you know, currently right now we're in a country where the average student uses about 48 distinct EdTech tools throughout the year. And this is also happening at a time where student mental health, feelings of isolation, and like just general student needs around academics and social emotional needs are growing like all the time. And then

You also have educators in schools. They're overwhelmed. They're trying to figure out individualized supports, either in person with students or using technology. But yet our scores, our outcomes are going down. So all of this is happening at a time where every part of the system is very overwhelmed. And I think at the time with fair schools, I was seeing all of this happening. And I was remembering back to what it was like for me when I was a principal.

I also felt that way. It might've been 48, you know, distinct ed tech tools, but I was also in a situation where we had a lot of different products that we were using. And I was trying to look at all of our systems, also be in classrooms, be with teachers, be with students, be with families, you know, talk to stakeholders and understand what was happening and put all of that in like a doc or something, just to understand like, okay, this is a full picture of my school. What do I do next? And.

That was really challenging for me. And I remember wanting to make sure I understood all of those pieces because I had learned the importance of mixed methods data analysis when I was in grad school and how important it is to not just look at someone as a number, but to really understand them holistically in order to make change. But doing that on a scale for 650 students even then was hard. And I continued to see that after I left.

the school-based role and was working with principals and had at first developed a non-AI version of that where my team, worked with a few researchers on my team at Fair Schools where they would help me create like these mixed methods, analyses for every school that we worked with. And then we would use that as the launching point for our improvement work. And always found that that was a much

Seth Fleischauer (05:45.633)
Hmm.

Stephanie Frenel (05:47.904)
more effective way of supporting both the systems around climate and academics and also family engagement. And so as AI was coming out, it was becoming more prevalent and popular. I stepped back from the work and said, hmm, are there ways in which principals want AI to help them? And when I sent out a survey and talked to like 100 principals across the US, they named

Like over and over again, know, data analysis would be really helpful, you know, and they might not have, have said like, want to mix something that does mix methods and is based on like these research practices. But, you know, I think knowing my work within fair schools that supported that, and then also what they were asking for, that's really where school ops AI came from. and so that's why, you know, it's a platform that aggregates both quantitative and qualitative data.

Seth Fleischauer (06:18.741)
Hmm.

Seth Fleischauer (06:43.765)
So there's a story there of actually going out and talking to people. I'm curious as to how you got a hundred principles you didn't know to talk to you. you could talk about that offline. but, the, and then this story of your own personal experience, the success that you found with mixed methods, data analysis, which I'm assuming I haven't heard that exact term, but I'm assuming from the context that that is, a combination of qualitative and quantitative data put together into a more holistic.

analysis, breeding, deeper understanding of the students, as you say, extremely hard to do at scale. is that basically the model that you've built the school ops AI program off of and, and w what is it about technology that allows that to be done at scale that you wouldn't be able to do with what you, what you had done previously, all doing it in person. cause I guess implicit to that question is.

don't we want human eyes on all this stuff? so, so what is the AI doing that is increasing efficiency, but still making sure that the quality is there, that the human is there to make sure, to, ensure that the proper steps are being taken.

Stephanie Frenel (07:57.228)
Yeah. So the AI is acting, you mainly in the background to support this at scale. So some of this is like AI specific. Some of it is also just statistics, you know, at a really high level and like being able to do the data science piece around it. And so essentially this information that we're getting from a school data quantitative and qualitative is coming into a secure

database. And then from there, the different codes and the algorithms that we've created, so still more still just like some of the math side is therefore working with the AI to look at this data and parse it in certain ways that allows for these insights to come together. So if it's more on the qualitative side, that say the school has specific rubrics that they're using, and so therefore there's coding that's within that.

we're able to build that into the model so that you're able to produce something that aligns with the different codes and different analyses and insights they would want. And then on the quantitative side, also doing that at the same time and being able to parse out that information. And then what the user sees is they're asking a question like, how are my fifth graders doing in math and why? So they might get the how are my.

fifth graders doing in math from a basic visualization or just like a regurgitation of information. But then as they're trying to understand the why piece, it will go to a deeper level. So maybe there's specific skills or standards within the math quantitative side that is missing. But then also maybe there's some key information about how the students are feeling about school.

or the relationship with their teacher or how they're collaborating with each other that can also be surfaced at that time. And so the human piece is still there. There's a reason why this tool is built to be collaborative. So any analyses or insight, you would want to do this within a meeting with a principal and their teachers or.

Stephanie Frenel (10:14.296)
you know, a teacher meeting where they're all kind of together looking at this, making decisions so that everyone is thinking through like the information they're getting, continuing to converse with the chat bot to get to certain deeper insights and levels so that there's always this human element within it. And then this is a tool that we've tested. So we've tested it for several months with users to really understand like what's working, what isn't, you know, how.

are the insights that they're getting so that we too are also involved in this process and are able to really support the level of accuracy.

Seth Fleischauer (10:52.019)
So is it, is it fair to say that that same process that you went through as a principal where you were doing mixed methods, data analysis, combing through all of the different assignments, maybe picking out certain words or phrases or, evidence of ability and putting that all into, some kind of.

Report or document where this is the, my description of how this student is doing that that process is essentially what the AI is doing, but doing it more quickly with more data. and then presenting that to the educator. And at that point, the educator can have like a, a check moment where they're like, does, this consistent with how I'm feeling about my student? is that essentially how that is working and functioning in the background?

Stephanie Frenel (11:43.117)
Yes.

Seth Fleischauer (11:43.85)
Okay. So then the question is, how does the quantitative data start working together with that? Because when, in my personal use of AI, I also use it for some of that kind of stuff, right? Like I find like it's, generally LLMs are pretty good at like, I have this gigantic body of text, you know, they're called large language models, right? Like I have this gigantic body of text. would like you to pull out.

Like what the main threads are that you're seeing in this body of text. I've been pretty darn satisfied with how LLMs can do that just for me and my own personal work. I've been far less satisfied with how AI can work with quantitative data. Just like getting stuff.

into a spreadsheet properly or even just doing math, right? Like it can, can in a system, like math is a system that's there's always one correct answer. And LLMs are like, let's be creative a little bit and mess with stuff, right? Like, make it sound better. so I'm wondering like, how does your quantitative data then come in and meet the qualitative data to tell a better story about the student?

Stephanie Frenel (13:05.174)
Yeah, no, that's a really good question. So when you think about your open large language model that we are using like a chat GPT or something like that, you know, again, it's trying to give you the best statistical answer and it's also usually trying to please you, especially if you're using, you know, chat GPT or Gemini. Yeah, exactly. So there's a level of there where there's

Seth Fleischauer (13:25.501)
so sycophantic, yeah.

Stephanie Frenel (13:31.298)
there there's going to be more error as a result, even if like LLMs in general are pretty accurate, there's going to be that because it has a motive per se, you know, not trying to give it human characteristics, but the engineers on the other side have given it that. So for us, it's a similar process where in our not, you know, not to go too much into jargon, but like within our prompt engineering, we have given our, you know, AI a very specific, you know, persona.

Seth Fleischauer (13:41.812)
You

Seth Fleischauer (13:46.763)
Yeah.

Stephanie Frenel (14:00.652)
what it's supposed to do, what it's not supposed to do, areas where it needs to ask for clarifying questions, areas where it needs to give information in this way, examples of what that looks like. Because we're FERPA compliant, we're following all the data privacy laws, we're not actually holding onto any of the data that schools give us, at least within our LLM.

So therefore we have to go in and tell it exactly what to do because it's never going to be able to learn from its interactions with our users.

Seth Fleischauer (14:38.177)
Hmm. So how, how do you know that it, that your FERPA compliant? think that's a weird, weird way to ask that. But like, how do you, how do you know? Right. Like I, I know, like, I know obviously you have a plan. How, how do you know that that plan is working? Like, how do you know the data is not getting through the cracks?

Stephanie Frenel (14:59.5)
Yeah. So that comes to my team. So the tool was built and is continuing to be monitored by someone who has a very strong data security privacy background. That was really key for me. I spent a lot of time reading FERPA, reading all of the different laws, looking at examples of what it looks like to make sure that our architecture aligned with, whether it's

the data privacy law federally, but also every state has their own as well. So like just really getting deep into that work so that when we developed our initial architecture, we were being really mindful of that. And when we chose the different tools that we're using, that we were also like being very mindful of how that looks. Because I know from us what we're seeing with our generative AI.

A lot of it is, like you said, you have a conversation, it's learning from you, but there are also models out there that do something, it's called a rag, but essentially it's reading it. So it's like looking at it and then using that to just like repeat that information, but it's not taking it in. So that's kind of a layman's way of trying to explain this process, but that's essentially what's happening.

Seth Fleischauer (16:20.289)
Got it. And so that leads to another question that I have about how exactly school ops AI works, where I just had someone on the podcast, she's a therapist and she was talking about an AI scribe and how, you know, same thing, compliant data protection. But then there's also some learning about this particular case that's happening between the AI bot and the therapist and the patient.

So that when they're in session with them, they're listening and then the student later can, access the AI bot and can gain or be reminded of insights that were found during the session itself. So each successive session, each successive engagement with the AI that the therapist and the patient are doing is teaching them more and more about this particular case and ostensibly making.

the bot more and more able to support the student when the therapist can't be there. I'm wondering how much of school ops AI has, insight building that, is based on multiple bodies of data that are fed to the program throughout time. Right? So like you could think about it, like from a school's level, you talked about insights that principals are getting. This is the original problem you're trying to solve is that.

Principals are drowning in data. How do they make sense of this? How do they know what to act upon? there's a snapshot of like what this data means today, but then there's also how this data has evolved over time, how this data comp compares to a data set that measured something similar, but different, how much is school ops learning about the school and does it improve over time? Or are you trying to keep it on that snapshot level so that people can make.

decisions based on the moment.

Stephanie Frenel (18:19.414)
Yeah, so also another great question.

Stephanie Frenel (18:25.87)
trying to think of the best way to respond to it. So there's actually like two houses that are happening at once. And there's one that is a secure database. The LLM is not in that secure database. And that's where a school's data lives, right? And so to your point around it having to remember certain information, it doesn't necessarily need to do that because that information is housed somewhere.

There's already like visualizations that are created that don't need AI for that, right? So if you're tracking historical information over time, like a graph or something like that, that AI is not touching that at all within our program. And then you have the AI when it's being called upon by the user. The AI is going, like I said, reading it and then producing the information based off of it. And it's doing in a way where it's not actually taking

any of the personal identifiable information. It's also doing it in a way where, you know, once the, that session and information is done, like we have controls over how long it remembers something, you know, remember in terms of memory. and so we've set that so that it's relatively limited to your point because we want to make sure it's not holding onto, information over time. And so.

as let's say that someone's generating a report, that report is there, that report is not going to live in the AI space. It lives in that database space. So then on our end, we're going in and saying, here's again, what you need to know, here's the information, here's all you need to know about this rubric or competency or this program or the standard, but all of that is just public.

Seth Fleischauer (20:00.449)
Hmm.

Stephanie Frenel (20:20.096)
information, right? It's not information that is specific to any student, any teacher. So the two kind of live very separately, but we've just created a model where it can speak to it when it needs to. But if let's say someone said, I don't want to use this anymore. We would lose all that information in our database and also the LLM therefore would not have it anymore.

Seth Fleischauer (20:25.483)
Sure.

Seth Fleischauer (20:42.753)
this has been my favorite kind of interview because I asked you one question at the top and I have not looked at my notes since then, because I just keep asking more questions and you keep having great answers for them. Clearly you've thought about all this very deeply. You have a great team in place. one of the other things I'm curious about, and then maybe I'll look at my notes again is, is like what you're testing right now. Like I know that you

have a couple of pilots running. and I I'm really curious, like you talked about accuracy. it sounds like that might be one of the things that you're testing. I'm curious, like what are you testing and how are you going to know if, if you're successful or not?

Stephanie Frenel (21:24.054)
Yeah. So for us, there's three key questions and a lot of this comes back to me being an educator, you know, coming from the space. I don't ever want to build something that isn't going to have a positive impact on schools. So, you know, that's a key part of it. But our first question for our direct users, which are, you know, the adults within schools and districts. So do users find the insights valuable? Is our first question that we're asking.

sorry, I hear, I don't, can you hear that, Siren? Okay. I'm gonna, I'll redo. Okay. Yeah.

Seth Fleischauer (21:56.614)
yeah, I do hear that a little bit. it it was okay. Yeah, let's, let's, yeah, that's cool. I think it's faded Chicago. Yeah, no, all good. Thank you. and actually, well, we're stopped. going to close my door. It's my dog.

Stephanie Frenel (22:06.442)
I know I was like, can I, should I keep talking through that? no.

Yeah.

Stephanie Frenel (22:22.67)
Oh, I see what you're doing.

Seth Fleischauer (22:25.087)
He's a good puppy. It's time for him to go out again. all right, cool. you ready to start over?

Stephanie Frenel (22:30.734)
Yeah.

So for us and for myself as an educator, as someone who cares deeply about schools and doesn't want to create anything unless it's having a positive impact on teachers and students, it was very important that we are thinking about the impact of the tool. So there's three key questions that we're asking as we're measuring the pilots. So the first one, do users feel that the insights they're getting

are valuable. Like are they valuable for them to get that information from the AI? Is it valuable to see the visualizations that they're getting? So that's like the first one. The second one is are they using the insights, right? It might tell you something, but if they're not actually, you know, using it, thinking about it, talking to their team about it, it's also not helpful. And then the third one is based on that, what is the impact?

Seth Fleischauer (23:23.937)
Hmm.

Stephanie Frenel (23:31.468)
on teachers and students. Are teachers doing things differently within their classrooms, are students therefore having a better school experience, learning more as a result. So we do have a researcher on our team that is supporting that work. So she meets with our pilot users and asks questions related to those three.

Seth Fleischauer (23:56.246)
Got it. Yeah. It's sound that sounds, that sounds complicated, like, like that last part, right? Because it's like, I mean, schools are so dynamic. There's so much going on. Like, are there specific pieces of impact that you're looking for? Are you just trying to tie them to the things that you were talking about? Like if, if it turns out that a teacher is using these insights to try to improve their classroom management.

Uh, practices that then you're looking for specific classroom management data, or are there like other sort of bottom line, like test score kind of stuff that you're trying to look at or all of it? Like, do you, what are you looking for?

Stephanie Frenel (24:34.06)
Yeah. And, you know, for the record, like we're not quite yet at like the randomized control trial level of this, you know, one day I would love to do that, you know, to be able to really say like, you know, here's the impact we have. So right now it's more anecdotal. is based on whatever metrics, you know, they're setting. like in your example, behavior, are they looking at behavior incidences therefore for a particular student or a particular class as a result?

of the information that they're getting. So, yeah.

Seth Fleischauer (25:06.945)
Got it. So yeah, it seems like there's all these different levels of insights from the teacher level to the administrator level in between that is the level of coach. And that's something that when I, again, referencing this therapy lab episode with the scribe, I was really intrigued by the idea of being able to use scribes for coaching where the dialogue between the, um, the.

coach and the teacher would be able to be fed back into some kind of system where it's learning more and providing insights. And then the coach wouldn't necessarily need to be there all the time. The teacher could engage with it. I, I'm not saying that that's exactly what you guys do, but I am curious about like the coach within this model. like how is school ops AI, supporting that specifically because we know.

that coaching is like the best way to professionally develop people. you know, all these insights are great, but that particular relationship is critical. So what, what are you guys doing to support that role?

Stephanie Frenel (26:13.09)
Yeah, so I think there's still always going to be an element of...

coaching that is human, right? So in terms of teacher development, at least in the work I've done in the past and a lot of the coaching work I've done or have studied, in order to get something to really be implemented into the classroom, sometimes you'll have a teacher who is able to learn something once, get a recommendation, implement it flawlessly. That is not the majority of people out there. That's just...

not the majority of learners out there, right? Where they get something once and then are able to do it perfectly. And so I think for us, it would be the assistant to the actual human coach that is there. So having the insights to see, here are some suggested areas where, and ways and strategies that you can support students based on the data that we're seeing here, or some insights around like what some of the root causes are for why a

particular situation is happening. And then from there, the coach saying, here's how I can implement or work with the team based on that information, so that there is a lot more that gets taken in and implemented into the classroom.

Seth Fleischauer (27:37.482)
Excellent. And you mentioned also SEL data, like social emotional data, either of the students and, or of the teachers. I'm not sure. obviously that is something that you can measure quantitatively if, with just mood scores and things like that. but qualitatively, I guess, assuming that people actually know how they feel, which is a whole other thing of like, let's like teach them how to know how they feel. but.

How, how does that all work within this system? Cause it, it feels like such a great potential to combine that kind of information with all of this classical academic insights to really tell a story. but I'm, curious how it, how it works within your system.

Stephanie Frenel (28:28.556)
Yeah, so are you asking how it works in terms of how we take in qualitative data or how it works in terms of the measurement side of things?

Seth Fleischauer (28:41.069)
I guess, yeah. Take, take me through the journey of like measurements to collection to insight, like what types and, and, and specifically how it's interacting with academic data. Like, is it just part of the picture? Is it working in the exact same way that everything else is? And it's just another part of the story.

Stephanie Frenel (29:01.344)
It's definitely just another part of the story. In terms of the metrics, we work with the school or the community or the district to figure out what that looks like. I think for us, we want to make sure there's a clear line between we're going to recommend this framework or we're going to recommend you use this model because then that kind of puts us in a place where we need to be the ones who are evaluating these tools.

knowing exactly what's out there. And I also think there's like a cultural context. Some tools work better in certain cultures, certain frameworks, right, than others. So I think for us, that's not necessarily a place we want to be in and where we're recommending things. So ultimately, if a district or a community has a framework, a value set they already want, they've already developed, and then they're thinking about or have already developed

different rubrics and metrics within that. That's what we take in is that information. And so then we tell our model, here's the different pieces. Here's what they're using. Here's how you need to think about this. Here's how you need to think about it in addition and connection to the quantitative and academic data. And then again, giving it like examples of what that looks like. So then when those questions come up by the user, we're able to really.

give more accurate insights in that way.

Seth Fleischauer (30:30.655)
Got it. Well, I'm glad I asked, cause that highlighted yet another part of the process here, which is like the training of the model, right? Like, and it sounds like for any given school, you're going to go through some kind of onboarding process where you need to understand exactly what kind of data they're looking for, what are their programs or their outcomes? and have a, have a deeper understanding of that before you can do anything with the work, right? Yeah. Makes sense.

Stephanie Frenel (30:52.492)
Yeah, definitely.

Seth Fleischauer (30:54.969)
I'm going to try something I borrowed from an Adam Grant podcast and I'm to have a lightning round here. and so, it doesn't have to be like short answers, I guess shorter answers, to a couple of like rapid fire questions. So one of them is what is something that you are rethinking,

Stephanie Frenel (31:13.642)
And by the way, I love Adam Grant.

Seth Fleischauer (31:16.822)
Hard not to load.

Stephanie Frenel (31:20.27)
I am rethinking, I know it's supposed to be lightning, but I like it.

Seth Fleischauer (31:25.889)
It's shorter, shorter answers.

Stephanie Frenel (31:37.08)
I am rethinking the entry point for our tool. think before I was really focused on just the school leader. And I think I'm recognizing that if we want this to be a collaborative tool, then that means we also need to engage like more around like the value around it. So I think that's one piece that I've been

really rethinking and have changed that approach quite a bit. yeah.

Seth Fleischauer (32:11.937)
Huh? That's interesting. Yeah. It's hard to find that, that critical stakeholder when you're trying to introduce something to a school, when schools are such collaborative dynamic systems that involve so many stakeholders and it affects all of them. And you know, how exactly to start that conversation. That's cool. That's interesting. Thank you. from either a personal professional level, is there a piece of media that you'd like to recommend a book, a show, a movie, a podcast?

Seth Fleischauer (32:59.071)
you're muted now. Did you, you touched your phone. no. Okay, you're back. You're back. Yeah. Sorry. Told you not to, I told you not to touch your phone and then, and then I gave you a question where you had to touch your phone.

Stephanie Frenel (33:06.958)
Can you hear me? Okay. I know, was ready to look at my own face.

Stephanie Frenel (33:16.778)
It's okay. What I'm currently listening to right now that I listen to all the time and would recommend is the knowledge project. I feel like I just learn a lot about what successful people are thinking, you know, and how they kind of handled some of the challenges within the world and also just get a lot of really good advice. I know it's only supposed to be one, but also the diary of a CEO is a really good one too.

Seth Fleischauer (33:47.579)
Ooh, neither of those are on my list, but now I'm going to have to. thank you. do you have any questions for me?

Stephanie Frenel (33:56.236)
Yeah, what do you think is the hardest part of getting schools to invest in looking at qualitative data regularly?

Seth Fleischauer (34:11.513)
part. I mean, it's just how alluring numbers are, right? Like everybody wants to grab onto something. They want to be able to like, say, no, no, no, this is the truth, right? I guess, I guess that's it. It's like the nature of truth, right? Like, you can, you can spin words more easily than you can spin numbers. And of course you can spin numbers too. but people I think have more trust in numbers because they have this certainty to them.

And that trust leads to a version of the truth that can feel more objective. even if we know that like, did you, any assessor, right? How did you ask the question? What, what did the kid have for breakfast that day? Right? Like there's so many things that can go into like a number that that that's where my work I've, I've always been more drawn to qualitative data, but you have to have a trusting relationship.

in order to be able to communicate to what qualitative data in a way that's going to feel meaningful.

Stephanie Frenel (35:14.926)
It's really interesting. I was like, can I respond to that? Okay. Yeah, no, I think that aligns with me. It's a both and, it's desire for certainty and numbers and then an aversion to something that feels very hard to measure and understand at a glance. then, but and at the same time, we're seeing in research that a lot of, you know,

Seth Fleischauer (35:17.568)
Yeah, please.

Stephanie Frenel (35:43.444)
information is coming out that there's a need for these qualitative pieces in order to better understand what's going on within schools and just within different domains. So it's interesting how the human brain wants to work and then what reality actually is.

Seth Fleischauer (36:02.145)
Yeah, that's the whole other podcast. maybe it's the knowledge project. don't know. well, that question that you just asked leads me to my final question, which, I'm wondering if you can tell us a story of like an aha moment of a school experiencing some kind of insight, perhaps around the qualitative data or just around your product in general, like,

When did, was there that like golden moment of like, my gosh, yes, this is exactly what I've needed. And here it is right in front of

Stephanie Frenel (36:37.42)
Yeah, so there's one.

story that comes to mind just around this piece of data triangulation, mixed methods analysis and being able to look across. And I remember I was working with one school and they were struggling both academically and behaviorally and there was just a lot happening. So it's very unclear where to start, what the entry point is. And this was a secondary school. So I think that also made it.

challenging too because developmentally where students are, you know, can't gimmick your way into getting students invested. You know, a lot of the little things that we'd often see around like points and things like that were just not going to work. And so through our analysis, we were able to identify that there was a major gap in student agency. So from the empathy interview data,

and the teacher interview data, students just didn't feel invested, but the investment was coming from the agency of ownership of feeling like there was this school and the work of the school was connected to who they were. So in the work that followed, they really focused on this agency piece. giving students opportunities to have that agency have the

you know, the information that they're doing in the classroom be more relevant and meaningful and also outside of school. So, you know, talking to students about the types of activities they want to do, you know, to support the school, but also like extracurricularly. So through that, and it wasn't like an overnight thing, but like your, we saw over time such a big change in the academics of the school. Like they had double digit gains within a year and a half.

Stephanie Frenel (38:34.38)
They also behavior incidents like within one semester was halved, you know, and it's just because they were able to be way more precise with the next step. And it wasn't to what we were talking about before, just we need to do this in order to get our academic numbers up. was, there's actually a piece around like value and investment that we need to work on with our students. And if we

create and craft the right strategy, which you get from the qualitative data, we're gonna see a difference. So, yeah.

Seth Fleischauer (39:08.673)
I love that story because you know, it's so rare that we hear good news in education. know, like there's just so much like there's so much as you said up at the top, right? Like things are in crisis right now. But the story you just illustrated is of an of a person in a position of power who was able to recognize the importance of these underlying root causes for achievement in academics.

something that no child left behind kind of, you know, wiped out. And I hope that we're more and more pendulum swinging back in the other direction of understanding that like, sure, that is one part of the picture. But if we just focus on that single part of the picture, we are not going to be able to actually, I mean, look at the numbers. We're not going be able to actually have an impact, right? But if with tools like yours, you can integrate

all of this information into gaining a more holistic view of the child and be more precise about what the interventions are that are going to identify those root causes, then we can start to make positive change. I love it. It's great. Thank you. Thank you for the work that you're doing. Where can our listeners find your work?

Stephanie Frenel (40:29.1)
Yeah, so you can visit our website, schoolops.ai. You can find me, Stephanie Fronell on LinkedIn. It's probably the biggest platform where I'm putting information up. And then if you do go through the School Ops AI website, you can also contact me that way via email or join our newsletter.

Seth Fleischauer (40:54.089)
Awesome. Well, we'll have all of those links in the show notes. thank you so much again for being here. was excited for this conversation and I am, I'm still uplifted, because, of, how inspiring this work is and how it, how it, it, it completes a piece of the puzzle, right? Like, being in an area of work, within this industry that is so hard to assess, that is so much better spoken about qualitatively.

It is a, it is a relief to know that there is a tool like this that can help integrate all of that information into one, source of action. So thank you for the work that you're doing. thank you to our listeners. If this, episode inspired you, please share it with a friend. follow us, leave a rating or a review. this episode was written and produced by me, Seth Fleischauer. It is edited by Lucas Salazar. Thank you, Lucas. And remember that if we launch.

Yeah, I almost got there. And remember if you want, sorry, that's not how I sounded. And remember if you want to bring positive change to education, we must first make it mindful. See you next time.