TRASCRIZIONE 🔗
WEBVTT
1
00:00:01.820 --> 00:00:12.420
Welcome to Charter School Insider, lessons from the nation's top charter school operators, the only podcast dedicated to the behind-the-scenes business of running a successful charter school.
2
00:00:12.920 --> 00:00:24.160
Join us as we sit down with charter school leaders who share how they are overcoming the biggest operational hurdles facing their schools today. DJ, great to see you. Welcome to the show. Thanks for being here.
3
00:00:24.900 --> 00:00:38.220
Thanks, Daniel. Excited to be here. Let's dive right into it. Give me the, or give our listeners the quick background on your path founding Kindle Education.
4
00:00:38.280 --> 00:00:50.980
What, what problem were you and your co-founder trying to solve? Yeah. I think for me the, the biggest problem was that school is boring and not a place where- [laughs]...
5
00:00:51.040 --> 00:01:02.380
students have a lot of opportunity to explore things they're curious in or be creative or things like that. I had issues with kind of the traditional model of school as a kid. Um,
6
00:01:03.800 --> 00:01:13.880
didn't really like school and then, you know, had some experiences in college that allowed me to do things I was more interested in, in ways I was more interested in doing them,
7
00:01:14.940 --> 00:01:26.880
uh, and show my learning in ways I was more interested in showing it. And, um, and I kind of got, you know, really into the idea of how do we do this kind of learning,
8
00:01:28.180 --> 00:01:34.560
which has been done in a lot of private schools for a long time. Um, can get into why maybe later.
9
00:01:35.420 --> 00:01:52.720
Um, but how do we do this in urban public school settings and give, you know, the students that we work with the opportunity to be builders and thinkers and explorers and creators and, you know, have the skillset to go on to help build the world and change the world and not just,
10
00:01:53.660 --> 00:01:58.460
you know, be a cog within it. Love it. Yeah. It's a great, a, a great vision.
11
00:01:58.920 --> 00:02:10.340
And for those who aren't familiar with the organization, give us kind of the quick rundown on what does Kindle look like today, you know, schools, what population do you serve?
12
00:02:11.060 --> 00:02:21.940
Um, a- and then for you personally as CEO, what, what does your day-to-day look like? [laughs] Just to kind of provide some context before we get into our topic. Yeah. It's a good final question.
13
00:02:22.340 --> 00:02:33.120
Um- I know that's a loaded question, yeah. [laughs] So, uh, my wife and I, we, we set out to start Kindle in 2020. We went out to High Tech High. We did a new school creation fellowship out there.
14
00:02:34.080 --> 00:02:46.580
Uh, we were very fortunate to get a lot of support, uh, along the way from a number of organizations. Uh, and the school opened its doors in the fall of 2023. After a lot of work
15
00:02:48.260 --> 00:02:52.920
meeting with, you know, local elected officials and community leaders and
16
00:02:54.140 --> 00:03:05.900
curriculum and writing a charter, uh, we ended up being the first school approved in the state of New Jersey, uh, first charter school in five years, uh- Wow... independent, so. Very cool. Yeah.
17
00:03:05.960 --> 00:03:16.440
We had a lot of people who, who didn't think it would happen, and there were definitely times where we didn't think it would happen, but, uh- [laughs] I'm sure... but, but we, we made it, we made it somehow.
18
00:03:17.220 --> 00:03:29.180
And so right now we, uh, we've grown a grade each year, so right now we're a full s- we're six through nine. About 66 students in each grade. We try to do 22 a class at the most,
19
00:03:30.360 --> 00:03:43.880
and we do a lot of project-based learning, competency-based learning. We're doing a lot of AI integration, trying to support students in, you know, developing the skills that I was referencing in my first answer.
20
00:03:44.520 --> 00:03:58.280
Now, as for what I do, um, lately I've been doing a lot of vibe coding. [laughs] Perfect transition. Yeah, incredible. [laughs] I've been building a lot of apps, uh, which- Mm...
21
00:03:58.320 --> 00:04:08.280
you know, I suppose we'll, we'll get into. Um, but, you know, I coach my two school leaders. I have a high school director, I have a middle school director. Great. We do school walkthroughs.
22
00:04:09.220 --> 00:04:20.800
Um, I lead some professional development sometimes on things like habits of discussion, um, and other, uh, learning strate- project-based learning, um, observe classrooms, give feedback.
23
00:04:21.500 --> 00:04:30.840
So it's, you know, it's nice to be able to still be on the ground and, like, be in the, be in the weeds- Mm-hmm...
24
00:04:30.860 --> 00:04:47.740
and then to also be able to pull myself out and think about what systems and structures can be built that, like, makes the work on the ground, um, you know, stronger and easier and more efficient and more aligned with our vision.
25
00:04:48.400 --> 00:04:49.280
Yeah. Yeah.
26
00:04:49.360 --> 00:05:10.860
I, I, I think that last point, you know, so many folks resonate with it's, and come back to, which is how do we make the actual work of education easier, uh, and free up time for staff so that they can do what they do best, you know, working with students, coaching students, and, and supporting, you know, really excellent instruction.
27
00:05:11.560 --> 00:05:24.700
Um, and you mentioned you've been doing a lot of vibe coding, which is, uh, a perfect transition to our topic today, which is talking a little bit about practical ways schools are implementing AI.
28
00:05:24.760 --> 00:05:28.420
It's one of the reasons I was excited to, to have you on the podcast.
29
00:05:28.740 --> 00:05:46.840
Before we get into specifics, I wanted to ask you, um, you know, as a school leader, as a leader and, and coach of people and an organization, when do you feel like AI started to become a
Load full transcript
serious topic and not just
30
00:05:47.740 --> 00:05:58.340
something you're playing around with late at night, uh, or, or, you know, talking to ChatGPT about your problems like so many of us- [laughs]... like so many of us are finding ourselves doing?
31
00:05:58.440 --> 00:06:14.224
You know, like, when-- what was the spark moment for you to say, "Hey, we really need to start thinking about, you know, AI integration"? Um, given where the technology's at Yeah, I think that it has developed over time.
32
00:06:14.544 --> 00:06:25.724
I was using ChatGPT in the spring, uh, winter of 2023 right before the school opened up, and I was using it
33
00:06:26.744 --> 00:06:39.584
to build teacher rubrics that were aligned with our charter application and curriculum that was aligned with our charter application and, you know, giving it, uh, resources to work with.
34
00:06:39.824 --> 00:06:53.184
And, you know, at that point, it could do jobs like that decently. Mm-hmm. Um, but you know, as the technology-- and you know, I ki- I worked with some of our teachers to do, to do that as well at the time.
35
00:06:53.784 --> 00:07:09.664
As the technology progressed, we started getting more into, uh, you know, students working with it, students learning about it, building tools for teachers. I think over the last six to 12 months,
36
00:07:10.724 --> 00:07:25.504
we've really taken the next step, and we now have teachers building apps with Playlab that students work with. I've been building a lot of apps with Claude Code, uh, which we'll talk about.
37
00:07:26.164 --> 00:07:37.744
My wife, co-founder Katie, has been doing the same. Um, and now we're at a point where it's like, okay, we're building too much stuff. [laughs] Yeah. We need to pause. The genie's out of the bottle, yeah.
38
00:07:38.184 --> 00:07:47.904
And make sure that, uh... And we have systems to do this, but make sure all of this is quality and also that we're not like, you know, wasting a ton of time building things- Right, right...
39
00:07:47.924 --> 00:07:55.464
that ultimately get used like three times and then, uh, thrown on the shelf. Yeah. Yeah. Totally. I feel like that, uh,
40
00:07:57.164 --> 00:08:10.884
a lot of us can resonate with that last comment just around, you know, the, the proliferation of dashboards and applications and, and, uh, you know, data analysis. What are you all finding in terms of...
41
00:08:11.064 --> 00:08:22.684
I, I know this is an ever-evolving, you know, work, so take this question with a grain of salt, but as you and Katie and the team are thinking through, um,
42
00:08:23.604 --> 00:08:29.364
some of those questions you were saying, which is like, where is this relevant? Are there any principles or,
43
00:08:30.464 --> 00:08:46.204
you know, guardrails or things that are starting to guide your decisions on where we want to lean in with AI implementation versus where maybe we feel like let's not, you know, or let's put that off for now? Yeah.
44
00:08:46.424 --> 00:08:56.324
I, you know, I think we're, I think we're implementing it pretty broadly. I think, you know, we're building things for students to interact with. We're building things for teachers to interact with.
45
00:08:56.384 --> 00:09:08.084
We're building things for our ops staff. We're building things for leaders. So I think, I think, you know, the most careful you have to be is when you're building things for students. Yeah. And for us,
46
00:09:09.404 --> 00:09:17.404
when we're thinking about utilizing AI with students, we are less interested in apps that
47
00:09:18.504 --> 00:09:31.804
personalize for an individual student, where they sit in front of the computer and they answer questions, and then it changes what kind of questions they're being asked. We like things that facilitate collaboration.
48
00:09:31.884 --> 00:09:45.144
We like things that require students to choose what they're learning and kind of direct their own path because we think those metacognitive skills are important. We think those, uh, collaborative skills are important.
49
00:09:45.384 --> 00:09:56.244
And so we, you know, a lot of our partnerships, we have a partnership with an organization, Oko, which is an AI that facilitates math conversations between students. So a lot of what we're looking at is
50
00:09:57.264 --> 00:10:12.484
how we can use AI to facilitate the kind of learning we believe in- Mm-hmm... which is not like student in front of a screen doing procedural question after procedural question- Right... until they understand the topic.
51
00:10:12.784 --> 00:10:30.324
Um- Yeah... I think, you know, for me, and I think this is just kind of like a design principle, having an initial idea and getting feedback on it before really digging in is important. Mm-hmm. Um, and then getting
52
00:10:31.604 --> 00:10:42.004
feedback along the way. So, you know, that kind of allows, that allows me to, and I think other people, to decide whether or not it's worth continuing to build this thing,
53
00:10:43.004 --> 00:10:51.124
whereas if I just built the whole thing and then sent it out and only two people used it- Right... it would, [laughs] you know, I would've wasted a ton of time.
54
00:10:51.124 --> 00:10:57.084
[laughs] Um, those are the ones that, you know, I think come to mind most. Yeah.
55
00:10:57.424 --> 00:11:07.104
Yeah, I think that, I think that's super interesting, uh, just on the, the adoption side and, uh, you know, what are we building that can actually get implemented.
56
00:11:07.824 --> 00:11:18.964
Let's talk about some practical examples, 'cause I wanna explore this and make it real for, for listeners. And I, I've heard some really unique things in the market just in terms of how people are playing around with AI.
57
00:11:19.064 --> 00:11:33.364
You mentioned you've been using Claude Code a lot. Can we chat through a few actual use cases that are kind of active at Kindle today and, and sort of unpack, you know, what was
the problem you were trying to solve?
58
00:11:33.404 --> 00:11:46.204
What ha- what's been some of the, you know, the downstream impacts? Um, how does that sound? Cool. All right. Should I, uh, should I pull them up on screen? Oh, cool. Yeah. Would that be helpful? Let's do it. Yeah.
59
00:11:46.924 --> 00:11:52.724
Is there a place you wanna start? Why don't we start, you and I were talking offline about, um,
60
00:11:53.644 --> 00:12:09.964
I, I think y- you correct me if I'm not describing this well, but you had basically turned your employee handbook into a, a chatbot that was unique to Kindle for all your staff, and I, I thought this was such a great example because-
61
00:12:10.888 --> 00:12:18.968
Anyone that's worked at any organization knows deep down inside they have not read the employee handbook, [chuckles] and then they need that answer to that one question.
62
00:12:19.068 --> 00:12:30.208
So maybe we can start with that, that first example. Great. So this is, this is Johnny Storm- [laughs]... named after, um- Yeah, give us the name origin...
63
00:12:31.048 --> 00:12:48.568
well, the Human, the Human Torch in Fantastic Four, his name is Johnny Storm. So, uh, I am nothing without my fire metaphors. [laughs] And we try to, we try to incorporate one as, as often as possible. Um,
64
00:12:49.688 --> 00:12:59.388
but essentially what Johnny Storm is, is the hub has all of our handbooks and playbooks as you can see, including our teacher rubric. It has all of our forms,
65
00:13:00.548 --> 00:13:11.628
and then, you know, it kind of has a direction to reach out to our middle school director. We can do a manual search. But the AI, and this was all built with Claude Code- Okay, great...
66
00:13:11.668 --> 00:13:26.368
the AI embedded in it, and this is a Claude artifact rather, so it's a Claude artifact, um, is that you can ask the AI questions about anything that's contained within any of these documents. Mm-hmm.
67
00:13:26.408 --> 00:13:40.228
And so that could be anything from, you know, what is the blackout day, uh, or what is the call out sick procedure?
68
00:13:43.008 --> 00:13:53.568
And so the first one will take a little bit of time as Johnny warms up. But- [laughs]... if he's working- Yeah... what you should see is that it will list that out, and then it will reference- Oh, yeah...
69
00:13:53.608 --> 00:14:00.048
where in the, uh, handbook you can find that. Okay. Which you can see here. Very cool.
70
00:14:00.468 --> 00:14:16.508
And so they- So this essentially has every, you know, handbook, teacher policy, any document that used to or at most schools lives, like, in some Google Drive folder, um, and has maybe sat there for, for years and years.
71
00:14:16.588 --> 00:14:27.808
This has all of those documents uploaded on the back end, and then Claude is just querying those and pulling responses. Yep. And then you can link right to where it is in the handbook.
72
00:14:27.848 --> 00:14:37.248
So the handbook itself is written right in here. Oh, very cool. Okay. So if you're just listening, DJ just clicked on the answer, and it jumped to the specific section of the handbook.
73
00:14:37.328 --> 00:14:50.328
And we'll link to this, uh, so you can watch the video in, in the show notes, um, and, uh, and check it out directly if you're just listening on audio now. So what is the, um... So, like, taking a step back,
74
00:14:51.448 --> 00:15:05.868
this is super cool, and I love that you can just build this. What was the thought process you had around starting here as an example of s- you know, a system or a tool to build for employees?
75
00:15:05.948 --> 00:15:12.248
What was the, what was the original thought process? And then follow-up question, what's adoption been like?
76
00:15:12.948 --> 00:15:20.868
We had built a few things that our teachers got into the habit of using and, you know, maybe there'll be time to show some of those too.
77
00:15:21.508 --> 00:15:27.448
One of them was a data dashboard that pulled from different platforms into one location. Okay.
78
00:15:28.168 --> 00:15:37.728
And use of that was strong, so we had teachers who already were kind of getting into the mix of using apps that we were creating.
79
00:15:38.988 --> 00:15:58.148
The thought process behind this was, I think I was having a conversation with, uh, Katie, my wife, co-founder, who also oversees all of operations, and we were talking about how much time, uh, and, you know, quite frankly how annoying it is when someone comes and asks about a question- Right...
80
00:15:58.188 --> 00:16:05.928
that we've reviewed 100 times and is in the handbook- Yeah... and could easily be found. Well, maybe not easily be found, but with a Control + F could easily be found. Right.
81
00:16:06.948 --> 00:16:17.848
Um, and I was like, "You know, why don't, why don't we just build an app where we have all that stuff in it, and there's a bot that you could just ask questions about that stuff?" Um,
82
00:16:18.748 --> 00:16:30.448
and so that was, that was the origin. We then thought, well, what if we put, like, our teacher rubric in here, and we put strategies- Mm... for how to improve in some of these areas in here.
83
00:16:31.128 --> 00:16:41.108
And the bot was able to talk with people about being able to improve some of their teaching strategies around whatever areas they were struggling in. And so it can do that now too. Very cool.
84
00:16:41.228 --> 00:16:51.248
Um, and then one of the uses that we found that, like, we, we didn't expect, which is the cool thing about these, is that you could build something and then it could have a use case that you never even thought of.
85
00:16:52.028 --> 00:16:56.888
So when we are interviewing someone, we'll send them this, and we'll...
86
00:16:57.028 --> 00:17:09.228
You know, if we're interviewing someone for, like, a dean of culture position, we'll ask them to review our policies and then to come up with a strategy or a plan or something based on
those policies, uh, for our interview.
87
00:17:09.268 --> 00:17:15.748
And we actually, we have someone interviewing, uh, on Tuesday, and so we just, we just sent it out. Very cool.
88
00:17:16.548 --> 00:17:37.348
And, and have you found, um, that just going back to the, you know, or- original pain point of it's super frustrating when people come and ask a question that we've covered 900 times and it's in the, in the employee handbook, you know, a-and for all of our friends in HR, they're nodding along aggressively.
89
00:17:37.648 --> 00:17:51.388
Um, i- have you found that it's cut down significantly on staff questions or requests? Like, I... Is, is it working as intended on that front in terms of the stream of, of information? Yeah. Yeah.
90
00:17:51.928 --> 00:17:58.608
And, you know, a lot of that was marketing. Yeah. It's one of the reason why I, you know, do the funny names and stuff so it sticks with people.
91
00:17:59.428 --> 00:18:08.708
Um, but, you know, whenever I was talking, whenever we were talking about something in summer PD, we were like, "And if you have any questions about that, ask- Ask Johnny Storm... Johnny Storm." Oh, very good.
92
00:18:08.728 --> 00:18:22.300
And so, uh, when teachers would come to our operations managers, they'd ask, "Have you asked Johnny Storm?" That's great. Um, and so- Eventually, yeah. I think everyone just kind of got used to going here first- Right...
93
00:18:22.300 --> 00:18:30.440
and just getting a quick answer... I'm just gonna be told to ask Johnny Storm anyways. [laughs] Right. Yeah. Right. That's great. Love it. Super, super practical example.
94
00:18:30.560 --> 00:18:44.340
And then you, I presume you can have Claude pull analytics on usage of Johnny Storm over time, and, you know, how many folks are using Johnny Storm. What... Super in the weeds kind of practical question, where...
95
00:18:44.600 --> 00:18:56.440
Does Johnny Storm get, like, regularly emailed out? Or since it's a Claude artifact, the link can be shared with everyone, and they can all be given access. But what's the...
96
00:18:57.000 --> 00:19:06.720
Like, what's the distribution of Johnny Storm look like? Maybe thinking about new school year starting, are you baking it into, like, new teacher onboarding?
97
00:19:06.920 --> 00:19:20.640
Is it, you know, linked other places in your internet or things like that? I know sometimes people are very curious about, okay, very specifically, how did you get this into the hands of your users? Yeah, yeah.
98
00:19:21.080 --> 00:19:33.460
We use Slack, so it's pinned- Oh, nice... in our operations Slack. We also have-- We use a learning management system called Headrush, and we have a teacher module in Headrush. Okay, cool.
99
00:19:33.480 --> 00:19:46.400
So Johnny Storm is in there as well for them to click on. Um- Nice... and then at the beginning of the year, I would basically just, you know, plug him in every message I sent- Yeah... over the summer.
100
00:19:46.420 -->
00:19:59.740
[laughs] Yeah, it's... Yeah. In your in- So I'd be like, "Is the food here yet? Use Johnny Storm." Right. [laughs] Right. Hyperlinked everywhere. Okay. I love it. Um, okay, super cool. Love that example.
101
00:19:59.900 --> 00:20:10.920
Also, very replicable by others. Um, e-everybody's got a handbook and a set of policies and things that you could roll up into a, you know, your own version of Johnny Storm.
102
00:20:10.980 --> 00:20:18.380
You'll probably have to come up with your own name. You can't steal DJ's- [laughs]... team's name. Uh, let's, let's talk about a second use case.
103
00:20:18.580 --> 00:20:37.720
Um, I think you had mentioned another to me around just your use of data and being able to think more strategically about academic data, um, you know, staff data that allows you to look at trends, make more proactive, you know, intervention for coaching and things like that.
104
00:20:38.080 --> 00:20:45.620
Could you unpack this for us a little bit? What, what does that look like, and, um, you know, where did you all start and kind of where are you iterating today?
105
00:20:47.380 --> 00:20:59.800
So we have our students, uh, set goals on Monday and review them on Friday, and then we have these work row lead outcomes, and then we have a goals log.
106
00:21:00.520 --> 00:21:09.300
I'll show you the data for now because it's most aligned with the question you're asking. So this is-- we're looking at this as if we were a student, or as if we- Yeah.
107
00:21:09.380 --> 00:21:16.400
So right now we're looking at it as if we were a student. Okay, great. Okay. Got it. But it'll show how the data kind of works. So
108
00:21:17.840 --> 00:21:29.420
this is a demo student, but in the student's dashboard, it pulls all of their past due assignments from- Mm-hmm. Okay... their learning management system. It shows everything that's due this week.
109
00:21:30.620 --> 00:21:43.420
It has all of our learning outcomes, and we can click down to see what those outcomes are, and we can click down to see the tasks that have been assessed for any of those outcomes. Mm-hmm. Yep.
110
00:21:44.700 --> 00:21:59.060
We can see how we're doing in any of our classes. And so in math pr- I'm doing, uh, proportional reasoning and my reason quantitav- quantitatively and model. These are the tasks that have, uh, been assessed.
111
00:22:00.300 --> 00:22:16.600
I can see the note that my teacher left me. Oh, wow. And I can even draft an email that includes all the data, their comment,
112
00:22:17.580 --> 00:22:30.860
and then a section for the student to write their question or their comment, uh, reminding them to be respectful and kind- [laughs] Yeah... when they do. [laughs] Helpful reminder there. Yeah. So if you're- Yeah...
113
00:22:30.920 --> 00:22:43.860
if you're just listening, um, DJ's showing us you can click directly on a comment from a teacher. It'll pre-populate the email for you to your teacher with the context on the specific assignment and the comment.
114
00:22:43.940 --> 00:22:51.400
So it, you know, saves you the time of, of drafting that, allows you to jump right to the kind of meat and potatoes of what you're trying to get to.
115
00:22:52.000 --> 00:23:07.340
And all of this, DJ, you said all of this is data that's being fed from the LMS, or is this a combination of systems? It's being fed from the LMS, and it's being fed from our, uh, culture dashboard deans list. Got it.
116
00:23:07.400 --> 00:23:20.680
Okay. And- So deans list, and then Headrush, I think you said. Was it Headrush? Is the- Headrush, yep. Okay. Got it. Great. And we'll soon be integrating our, uh, observation and feedback, uh, data into it as well.
117
00:23:21.520 --> 00:23:31.620
We use a partner, Attuned. Got it. Okay. And so as a student, I can understand why this would be super helpful for me just to understand where am I at.
118
00:23:31.740 --> 00:23:46.880
Like, let me zoom out and see where I'm at across all my, all my classes. Are you all, like, coaching them to use this dashboard, and, and are they sitting down with a teacher or an advisor to talk through this?
119
00:23:47.220 --> 00:24:01.860
Tell us a little bit about the practical implementation. Yeah. So every student, we start every day with a class we call Spark. Okay. And so in our middle school, it's about 22 students and three teachers/tutors, admin.
120
00:24:02.660 --> 00:24:13.100
Um, we have a three adult to 22 ratio because, you know, for coaching, um, you want smaller groups. Mm-hmm. I mean, for everything you want smaller groups- [laughs] Right... but particularly for coaching.
121
00:24:14.160 --> 00:24:31.392
And so the way that this works is on Mondays, students will come in here, and they'll click this set goals- Okay... and it will show them each of these things To review. And then they can look through all of this.
122
00:24:32.292 --> 00:24:49.072
Yep. They'll choose what data they're setting a goal in response to. So maybe I'm doing past due tasks, and maybe I'm doing, I don't know, this one, this class.
123
00:24:50.092 --> 00:25:02.052
And then you can type in, you know, whatever your goal is. Right. This week I will- Committing to... to do this, I will- Sure... I will know I did it, and then I save it, and I go to the culture.
124
00:25:06.832 --> 00:25:17.952
And this is, you know, essentially the same. Mm-hmm. I choose the number. Right. You know, maybe I've gotten a lot of smolders, and I see I've gotten most of my smolders from Mr. Lindquist,
125
00:25:19.032 --> 00:25:30.192
so maybe that's what my goal is gonna be around. Mm-hmm. And then, uh, these are our, you know, what some people refer to as durable skill competencies,
126
00:25:31.231 --> 00:25:45.212
and so like set goals is one, prepare and organize, practice wellness. And these are also assessed on tasks over the course of any given week, month.
127
00:25:46.292 --> 00:25:57.892
And so I pick, you know, I can pick two that I wanna set a
goal around. So maybe manage time and take feedback, and I could say, you know, something along the... I will incorporate...
128
00:25:58.292 --> 00:26:10.432
Well, we want them to be specific, which is why I'm not trying to think of them- [laughs]... off the top of my head. Right. Then it has these goals. Oh, okay. So it spits out your summary of your goals. Great.
129
00:26:10.872 --> 00:26:26.272
Then I submit them, and then it keeps a log. Okay. And so this pushes directly into our teacher data dashboard, which is what we were talking about before. Hmm.
130
00:26:26.332 --> 00:26:38.292
So they have their Spark students, and they can look at all the goals their Spark students set. I see. Um, it also, this is available to parents, so parents can come in here and see all of this data. Interesting. Okay.
131
00:26:38.992 --> 00:26:53.452
And then the, on Friday, students would reflect on the goals. Should I make this mandatory for students to have to look through it before they set goals?
132
00:26:53.472 --> 00:27:10.832
[laughs] So they don't just skip the tutorial like everyone always does. And then I save my reflection, and then my reflection goes right in there alongside, uh, my goals. Right, with everything else, yeah. Very cool.
133
00:27:12.752 --> 00:27:27.292
And so your, so this is students and staff are both utilizing this Spark dashboard, and it's, it's keeping all the data kind of warehoused in one place to track student progress and goals over time.
134
00:27:27.792 --> 00:27:43.052
And then for you all as a leadership team, are you then looking at this data kind of at a macro level across the entire school to say, you know, how are we doing as an organization against our, you know, the things we're trying to move the needle on?
135
00:27:43.752 --> 00:27:50.712
So what I can see here is I can see my teachers, and I can see how many of their students are off track,
136
00:27:51.892 --> 00:28:02.792
the percentage of their students that are partially proficient, proficient, how many overdue grades the teacher has, how many past due tasks the students in the teacher's class has,
137
00:28:04.172 --> 00:28:12.832
their percent positive, the number of referrals they've given, the number of family contacts, um, and the number of students on their watch list.
138
00:28:13.572 --> 00:28:26.472
And then if I want, I can click into a teacher, and I can see everything that that teacher has to do. Wow. I could go to their Spark students, and it would show me their Spark students.
139
00:28:27.752 --> 00:28:41.612
And the, the teachers, the teachers have the same view available to them, so when they log in, they can see their dashboard, which has all this information. Right.
140
00:28:42.892 --> 00:28:57.572
And they can go and see their Spark students, which again will have to... And they can see what their Spark student's goal was for this week, their schoolwork, their culture, their work, their reflection.
141
00:28:58.612 --> 00:29:12.272
They can email the student or their family directly from the dashboard. They can look up previous goals that have been written. You know, this data is available for all the data we've been looking at. Right. Right.
142
00:29:12.732 --> 00:29:29.272
And so previously, you know, if, if I rewind the clock, a teacher would've had to go, you know, to your culture platform, to your LMS, to your, I don't know, maybe your SIS, wherever, like
143
00:29:30.232 --> 00:29:45.032
disparate systems to get all this data. Now it's all pulled into one system and presented maybe in a more aesthetically pleasing, uh, and, and kind of accessible manner. And how much of this is
144
00:29:46.172 --> 00:29:57.372
do you feel like is unlocking new things you didn't previously, you know, couldn't previously see trend lines on or didn't have the time to track? 'Cause I think that's one of the most interesting things
145
00:29:58.312 --> 00:30:10.272
is AI is uncovering, you know, it's so good with data and with text that it's uncovering things that previously it would've just taken you hours and hours and hours to try and do some sort of comparative analysis on.
146
00:30:10.332 --> 00:30:18.312
Are there any anecdotes there just in terms of, you know, observations or learnings you all have had, uh, as you've worked through this process?
147
00:30:18.932 --> 00:30:27.412
Yeah, I mean, there, a lot of the views, probably the majority of the views that I was showing you are not available in either dashboard. Hmm. Interesting.
148
00:30:27.452 --> 00:30:42.180
It's just now that I have the raw data, I could just say, "Hey, I wanna see this from the raw data." Um, the other thing that I forgot to show you is that, so we also have an AI chatbot embedded inside of this.
149
00:30:43.040 --> 00:30:55.740
And so I as a leader could say, "Which of my teachers have the most overdue grades that
150
00:30:57.280 --> 00:31:03.600
are at least 10 school days old?"
151
00:31:06.740 --> 00:31:20.280
And now this is gonna query the data for the specific question that I asked, and it's gonna return a response similar to what you saw with Johnny Storm, but now it's looking through our data. Hmm. Interesting.
152
00:31:21.100 --> 00:31:34.180
So s- so teachers can kind of have, you know, explore their own learning paths here just on, you know, my students, my classes, what observations do you have, data, how does my data compare to data across the organization?
153
00:31:34.980 --> 00:31:44.360
It's kind of all self-serve through the chatbot as well. Yeah. Awesome. So leaders can, teachers can, eventually students and families can.
154
00:31:44.520 --> 00:31:58.440
We're gonna embed a Playlab app into their version, and they'll be able to ask questions about their data. Super cool. Okay. So l- let's, uh, let's zoom, let's zoom out. This is fascinating, and I think, uh,
155
00:31:59.440 --> 00:32:07.440
clearly you guys are pretty far down the path here with, uh, with how you're, you know, implementing AI. What would you say...
156
00:32:08.120 --> 00:32:20.680
You know, there's probably someone listening to this or watching and seeing what you guys have built and thinking, "Oh man, that... They're like way-- They're on second base. We're still sitting in the dugout." Um,
157
00:32:21.680 --> 00:32:38.380
what, you know, if you were starting over or if a friend calls you and, you know, it's a fellow school leader and they said, "Hey DJ, I'd like to get started on leveraging AI a little, a little bit more," what, what would you tell that person?
158
00:32:38.680 --> 00:32:53.300
Any recommendations for, um, starting over or learning through this journey? So I have no coding experience. I, I couldn't code a single thing, um, with, with actual code.
159
00:32:54.220 --> 00:33:07.600
And I've been really into AI and talking with AI and, you know, just generally. But I really only started Vibe coding, you know, less than a year ago. Um,
160
00:33:08.620 --> 00:33:18.980
and the way that I learned how to do it was just talking with AI. So I'd have one AI up that was Vibe coding.
161
00:33:20.060 --> 00:33:30.260
I'd have another AI up that was answering questions about the Vibe coding, and I'd just say, "Hey, I wanna do this next." Mm-hmm. Like, "How do I do it?" Mm-hmm.
162
00:33:30.900 --> 00:33:37.700
And then it would tell me, and then I'd do some of it, and I'd say, "What am I doing wrong? What questions do you have?" Mm-hmm. And it's like
163
00:33:38.620 --> 00:33:50.820
you basically have a partner who is an expert at doing this thing who will answer any question you want- Right... for however long you want, with whatever challenge level you want. Yeah.
164
00:33:50.880 --> 00:34:00.540
And you could say, if you want, "Say that again, but on a third grade level." [chuckles] Yeah. [chuckles] You know? That's like my entire chat log. If you looked at my- [laughs]...
165
00:34:00.560 --> 00:34:13.440
Claude chat log, it's like, "No, no, treat me like I'm in second grade and try again." Yeah. So I mean, you know, I think, I think a lot of people see... And there's a lot to, to be nervous about with this technology,
166
00:34:14.820 --> 00:34:25.920
but this also basically gives people the example to learn almost anything, and you always have to verify it, right? And like code's good because you can see if it works or it doesn't work. Mm-hmm.
167
00:34:26.000 --> 00:34:38.720
But to learn anything with, you know, an expert at your fingertip twenty four/seven. Yeah. It's pretty incredible. Yeah. What are you excited about going forward?
168
00:34:38.800 --> 00:34:44.380
Like as you, you know, as you think about, okay, we've been able to do this, this, and this,
169
00:34:45.320 --> 00:34:54.839
uh, when you can't sleep at night and you're a- you're iterating on AI ideas, what, what are you thinking about, you know, as you look out over the next twelve to twenty-four months?
170
00:34:55.320 --> 00:35:07.060
Well, do you wanna see the
wildest thing we're, we're doing- Sure, yeah... that we launch in January? Heck yeah. [chuckles] So, um, where do I start? As long as there's no student data.
171
00:35:07.080 --> 00:35:19.180
[chuckles] There is no student data in this one. [both laughing] You gotta watch out for that FERPA. Okay. So now with-- We have a number of partners we use, and there's also a ton of open source curriculum. Okay.
172
00:35:19.200 --> 00:35:34.700
And now with these, uh, agents that, you know, like a dot or an agent that can just do work on its own, I've been having it find open source material and put it into folders aligned with our curriculum.
173
00:35:35.160 --> 00:35:51.200
And so this entire folder, and basically every folder for every subject, has been built by an AI agent going across the internet and looking for open source materials and putting them into the folder that aligns with the curriculum for any given module or unit.
174
00:35:51.820 --> 00:36:06.220
Huh. And this is just while you're sleeping. This is, yeah. This is whenever. While I'm sleeping. I then have something, a builder that turns it into what you see here- Okay...
175
00:36:06.880 --> 00:36:20.360
which is a learning module in our learning management system, Headrush, and it has Learn with the videos. So like the, the dot AI found this video and put it in. Right. And then the,
176
00:36:21.360 --> 00:36:27.400
I call it Gwyn, the thing that builds this. Gwyn then takes that video and puts it in this structure.
177
00:36:28.720 --> 00:36:43.996
And so you have a ton of stuff in here from opportunities to learn, to practice, to get remedial support, to do more advanced complex things, to build projects. And then it builds a converter guide
178
00:36:45.176 --> 00:37:01.916
that will convert all of this [horn honking] into an open world simulation for students to navigate, where they can access any task and do it through there. And what that looks like-
179
00:37:03.536 --> 00:37:20.796
Treat me like I'm a second grader, DJ. [laughs] Slow it down. [laughs] [keyboard clicking] Kindle Lands. Yeah.
180
00:37:23.136 --> 00:37:36.796
Wow. So you talk with the mayor, and the mayor tells you, "I finished it." But the mayor tells you you need to light the city, and then you do a little tutorial if you want. But really, you're this little character,
181
00:37:39.216 --> 00:37:53.116
you can go places, and if you go to the library, you can pull up those resources we were looking at. Yeah. Yeah. And it'll open it in Head Rush. Huh.
182
00:37:55.516 --> 00:38:09.796
And you can go to [keyboard clicking] the shipyard, and that's where all your practice and remedial support is. Mm-hmm. And then
183
00:38:11.796 --> 00:38:24.736
at each lighthouse in each town is your final proof for the topic. Mm-hmm. And then there are also, like, mini games within it.
184
00:38:26.116 --> 00:38:43.856
Um, so like this one requires students to walk across and hit a number when they were on certain parts of the
number line, and that's what lit up his lamp. Um. And so all of this was just created based on
185
00:38:45.236 --> 00:38:59.476
the various modules that, that are in the curriculum outline? Yeah. All of it was created based on my dot uploading all of those resources into a folder. Mm-hmm.
186
00:38:59.856 --> 00:39:11.816
My Gwyn agent turning those resources into that Head Rush module- Mm-hmm... that you saw. Yep. Along with a converter guide. Mm-hmm. And then this, Kindle Lands,
187
00:39:12.796 --> 00:39:27.836
reading all of that, and reading the converter guide, and turning it into this open world- This world... video game you essentially see right now. Fascinating. And what, um... 'cause I know people are gonna ask, what,
188
00:39:28.816 --> 00:39:46.496
what, uh, model are you using to do this? Or, you know, what- what's building what, I guess is the, like, the question. Yeah. Good question. So the dot is just ChatGPT dot. Okay. Yeah.
189
00:39:46.556 --> 00:39:55.776
Um, and if you have like a pro account, you can access that. Right. You can do that. I pay extra to have more, like- Sure... work memory. Yeah. Um, because obviously I'm using it a lot.
190
00:39:55.916 --> 00:40:10.056
The thing that turns it into a module is a Claude code project. Okay. Yep. And then the thing that takes it from being a module and turns it into what you see here is also a Claude code project. Got it. Okay.
191
00:40:11.076 --> 00:40:22.336
And whenever I'm using Claude, I try to use either, uh, Fable or Opus 5.5. Sure. Yeah. So the higher-end models. The higher-end models, yeah. More capabilities. It just make it easier, you know? Yeah.
192
00:40:22.356 --> 00:40:31.596
So we have like a school max account. Yep. So that we can, you know, I can build all these things without hitting limits. Right. Right. Okay. Very cool.
193
00:40:32.016 --> 00:40:41.216
Uh, well, this is a perfect place to end 'cause I feel like it, uh, gives us some aspirational, uh, creative juices for those who might be listening.
194
00:40:41.616 --> 00:40:51.616
DJ, thanks so much for sharing what you guys are up to, what you're building, and, uh, hopefully providing some, you know, some inspiration for others just to get started somewhere.
195
00:40:52.296 --> 00:41:02.575
Uh, any, any final resources or recommendations for listeners that you think we should put in the show notes as it relates to upskilling yourself with AI?
196
00:41:03.096 --> 00:41:13.176
Honestly, everything I learned was just from talking with AI and building, and I think, you know, for me and for a lot of people, I'm sure that'll be like the easiest way and starting small.
197
00:41:13.236 --> 00:41:23.536
The first thing I coded was this little thing, and I was just like, "Wow, like I can do this?" [laughs] And that's like everyone's reaction when they do it. Right, right, right. They're like, "Wow, I can do this?" Yeah.
198
00:41:23.616 --> 00:41:26.436
That's awesome. Great. And then it's like you just get better and better and better and better.
199
00:41:26.916 --> 00:41:39.156
And then if anyone wants to talk about this stuff or, you know, look through builds or anything, I, you know, I love to do so, and I'm happy to hop on a chat and, uh, and connect. Love it. Yeah.
200
00:41:39.316 --> 00:41:50.016
Find, find DJ on LinkedIn and, uh, uh, reach out if you wanna talk about this stuff further. And thank you again for taking the time. This was great. Thank you, Daniel. I really appreciate you having me.
201
00:41:50.816 --> 00:42:00.436
Um, I don't know if the virtual world was too much, but, uh [laughs] but all the other stuff is pretty good too. [laughs] Yeah. Yeah, yeah.
202
00:42:01.396 --> 00:42:10.916
You gotta, yeah, you gotta give some future-looking, uh, some future-looking sneak peeks. Yeah. Love it. All right. Sure. We'll wrap there. Cool. Have a good one.
203
00:42:10.956 --> 00:42:19.376
[upbeat music] Thanks for listening to Charter School Insider. This show is brought to you by BuyQ, the group purchasing organization specifically for charter schools.
204
00:42:19.796 --> 00:42:27.836
Wanna learn how to maximize your school's purchasing power? Join together with thousands of other charter schools by visiting buyq.org/signup.
205
00:42:28.336 --> 00:42:38.816
BuyQ's purchasing program is free and has no obligations or commitments required. If you enjoyed today's episode, do us a favor and subscribe on your podcast listening platform of choice.
206
00:42:39.236 --> 00:42:48.136
If you have a question or feedback for us, we'd love to hear from you. Send us an email at podcast@buyq.org. Thanks, and we hope you'll tune in for the next one.