S2.E2 | Jose Moreno | From Netflix to Neulight: What happens when you bet on yourself?

Thembi talks with Jose Moreno, Founder and CEO of Neulight, an ed-tech company leveraging AI to help K-12 educators support student success.

In this episode, we discuss:

✅ What it’s like to transition from streaming tech to education
✅ How AI can revolutionize student interventions
✅ The impact of leadership styles on company culture
✅ Bias in AI and strategies to mitigate it
✅ Why sitting all day might be worse than smoking (seriously!) 

Plus, Jose breaks down his vision for Neulight, a company tackling real problems in education while challenging the traditional startup hustle. 

Neulight Website – Try their free version of IntelliTier

Connect with Jose Moreno on LinkedIn

Want to be a guest on KeyBARD? Send Thembi a message on PodMatch: https://www.podmatch.com/hostdetailpreview/1740803399472257afce75768

KeyBARD is produced, written, and hosted by Thembi Duncan.
Theme music by Sycho Sid.

Listen and Connect:

• Spotify: Follow, Rate, and Share
• Apple Podcasts: Subscribe, Review, and Rate
• Instagram: @Keybard_IG
• Facebook: The KeyBARD Podcast

Transcript

Jose Moreno: We do have strategies that we put in place to make the AI aware that it could be biased. We have things that we put in place to say, "Don't do that."

Thembi: Hello, hello, and welcome to KeyBARD. I'm Thembi, bringing you compelling conversations and content about technology, education, and the arts. 

Imagine this: you are working for one of the biggest streaming platforms on the planet, and you help build a feature that millions of people use every single day without thinking twice. 

Then later you work on a new project that brings in billions of dollars for the company, even though it makes subscribers a little cranky (more on that later). But in that, you come to understand the incredible value that you bring to the company. So you walk away and all the big kids are calling you over to play for their team next, but nope, you bet on yourself. 

That's exactly what Jose Moreno did. He went from Microsoft to Netflix to founding Neulight, an EdTech company that uses AI to help teachers better support their students. 

So what's it like to shift from the streaming world to the classroom? How can AI be leveraged to create more effective student interventions? IPO or no, IPO? We get into all of that plus leadership styles, bias in AI, and even why sitting at your desk all day might be worse than smoking. 

Stay right here. Up next is my conversation with Jose Moreno.

Jose Moreno graduated from Harvey Mudd College and began his tech career at Microsoft, working on services for over 500 million users. At Netflix, he contributed to features like Skip Credits and technically led the account-sharing initiative, helping Netflix add over 40 million new subscribers over the following 12 months. 

As founder and CEO of Neulight, Jose now focuses on transforming education through technology. His company's product, IntelliTier, leverages AI to empower educators to provide personalized student support, aiming to make a positive impact on education. Welcome Jose.

Jose Moreno: Thank you. Thank you for having me. I'm really excited to be here.

Thembi: Yeah, it's good to have you. So let's jump in. So, okay, you started your career at Microsoft and then you moved on to Netflix and now you run your own startup called Neulight. So I want to know what the company culture was like at Microsoft and Netflix and what philosophies have you brought from there into your own company culture?

Jose Moreno: Yeah, so I spent about five and a half years at Microsoft my first few years out of college, and then the last 10 at Netflix. The reason that I lasted so long at Netflix was because I really love the culture that they built. I really believe that a lot of the success that Netflix has seen has come from it. So I'll talk a lot about the Netflix culture that I'm bringing into the new company, but for Microsoft, I think it's a much larger corporation. There was always a stereotype that Microsoft was slow and heavy and reluctant to try out new technologies. 

For my first four years at Microsoft, I didn't see that at all. I was part of a new organization that was focused on building cloud platforms, and so everything was new and we were moving really quick. And so when people said Microsoft, I was like, I don't get it. I don't see it. And then for my last year, year and a half, I moved over to Exchange to help them out with their cloud solutions, and that's where I really started seeing some of that slow stereotypes from Microsoft. And so I think you get different experiences depending on where you are at Microsoft.

Thembi: Okay, so because it's so huge.

Jose Moreno: Yeah, because it's just so huge. Netflix really believed in empowering the employees, the actual people on the ground that are doing the work, and so they had a huge culture around hiring the best of the best, and then they saw it a sports team where a lot of these other tech companies are calling it families and come work for us. We're a family, we'll give you food, you can do your dry cleaning here, you can exercise. 

I never actually bought into that messaging where Netflix actually took a lot more of a straight to the point sort of take, which was come here, do your work, do amazing work, and then go home. And if you're a part of that, they pay really well and if it's not working out between you and Netflix, then separate parties can kind of move on. So it's a little bit more cutthroat, but it was really focused on having you and your coworkers all being super reliable people so that you can just move forward and push things ahead.

Thembi: So it seems like the best then would rise to the top in that kind of environment.

Jose Moreno: And they were quick to let people go with a really generous severance package and just say, things aren't working out. We're looking for a different type of person.

Thembi: I love that approach. It's just get to the point and then if it's not working, let's keep going.

Jose Moreno: There is no tracking of how much time you're working or anything like that. It was: deliver value, deliver features, and then go home. And I really love that. I think that's how it should be, especially in this American economy. Let's be real. That's what they're asking from us, right -- to actually deliver value.

Thembi: Yeah.

Jose Moreno: And so, if you're going to be honest about it, then just be honest about it.

Thembi: Okay, so lemme get this straight. You are a contributor to the skip credits feature of Netflix, which is really, do you think about that sometimes of how impactful that is? Millions and millions and millions of people have. Sometimes I think of it as sort of a small thing, but I use it every single time I'm watching something on Netflix.

Jose Moreno: Every time. Every time.

Thembi: And so...

Jose Moreno: Sometimes you don't even think about it.

Thembi: No, no. But if it wasn't there, I think that would be bad. So do you think about that level of impact?

Jose Moreno: I do, and this is something that I've been self-reflecting on. I have to be more willing to look back at my accomplishments and just be super grateful for what I've done and appreciative, but I'm the type of person where I accomplish something and then I'm like, all right, what's next? What's next? What's next? And so I'm always kind of forward looking when I turn around and actually look back at everything that I've done, then I'm like, yeah, that's a pretty successful career, but I don't know, it's always about the next day. 

That is something that I'm trying to work on in myself because I feel like you can never truly be happy if you're always thinking about the future and what you're going to do next. You have to just take a step back and actually be appreciative of the stuff that you've done. But thank you for saying that.

Thembi:

Yeah, it's a big deal. It benefits the end user, right? But what's interesting is that you also contributed to something that hugely benefited Netflix versus the end user, which is the account sharing thing that everybody was grumbling about, right? So did that give you a sense of your value? Did you then think, okay, I have a tremendous value.

Jose Moreno: Yeah...

Thembi: Let me think about this and see how I can do something for myself?

Jose Moreno: Definitely. After that project, I think I did start looking at my value a little bit more. That project brought in somewhere between three to 5 billion for Netflix over the following 12 months.

Thembi: Oh my gosh.

Jose Moreno: It's some pretty significant impact, and I think I knew that once that got delivered and everybody started to see how impactful that was, then my personal free market value went up significantly. And so I think that's where I started evaluating my options of what I wanted to do next. Was it staying at Netflix? Was it looking at some of these AI companies that were starting up? Because I do believe that generative AI and ChatGPT is providing a significant amount of value that did not exist before, and so I had to look at those companies to see if I wanted to join one of them. 

The problem with joining any sort of startup is that you're really dependent on the founders and if they're going to be able to make it successful or not. And I didn't really know any of the founders, and so it was hard for me to trust that if I went there, it was going to be worth my time. And then partly of appreciating my value a bit, I started to think maybe it's worth for me to start something up and have that impact that I delivered for Netflix, that I delivered for Microsoft and start delivering it for our own company. And so that's kind of where Neulight started up. 

Thembi: I love that -- bet on yourself. I think that's fantastic.

Jose Moreno: Bet on myself. That's exactly what it --

Thembi: Bet on yourself --

Jose Moreno: Putting the money on the table that I actually will be able to make this successful. And so that's where we're at.

Thembi: Yeah, that's awesome. Okay, so this is an EdTech company that, I mean, would you call it an EdTech company first off?

Jose Moreno: Yeah, absolutely.

Thembi: Okay. So educational institutions usually get criticized for being slow to adopt, right? Innovation, new ideas, and you have ed tech tools that are helping to streamline the process of evaluating behavioral issues in schools. Am I describing that well?

Jose Moreno: Yeah, that's correct.

Thembi: Okay. And so how do you see your tools that you're creating as something that can maybe help the education field pick up the pace and speed up? We're still working with paradigms from the early 20th Century in education. How are you going to help us out with this?

Jose Moreno: Yeah, so the reason that we started focusing on education was because we kept hearing this feedback about how terrible the software is that parents or administrators or teachers have to deal with on a day-to-day basis. Perfect. I know how to build amazing products. I have a network that I can help bring in to build amazing products. 

And so just from that standpoint, from the competitive nature, I thought we had a leg up on the competition. We just needed to get started. The second part of it is bringing in some of this new technology to move things forward and provide value that's never been provided before. So personalizing or providing personalized recommendations for interventions for students based off of a whole child view. I think that's what educators have been trying to do. 

But with the MTSS or PBIS frameworks, it seems like it's been a little bit more challenging to get people to wrap their brains around exactly how to move forward with it. And so that's where I thought that AI can provide some significant value. So the way that we're approaching this, don't get me wrong, we've gotten pushback on AI quite a bit,

Thembi: I'm sure...

Jose Moreno: Because it's new and people don't really know how to get the value out of it, and that's understandable. So the way that we're approaching it is we look for the people that are willing to push forward on innovation first. And we've gotten a few districts to start using in IntelliTier, which is great. We're in the middle of our sales period and things are off to a good start on finding more districts to get people to jump onto in IntelliTier. 

But what we want to get to is to start showing the impact of what in IntelliTier can have on the schools and districts in terms of improving student performance to then make that the focus of the sales. Because at the end of the day, what I've seen is that everybody is passionate about moving students forward and having them perform better. Everybody loves that concept. And so if you start showing that this product can have real results in terms of improving those outcomes, then it becomes easier to have the conversation with people that are a little bit more hesitant on the AI side because they want to improve those metrics as well.

Thembi: So as you go after this proof of concept that you can demonstrate to others to get them on board, what kinds of schools or districts are you finding are jumping on board? You said folks who are more innovative, but are any of these title one schools or districts or

Jose Moreno: We're in a virtual district, fully virtual district, and then we're in a district in Montana, which I thought was pretty interesting. They were one of the first ones to knock on our doors as opposed to us knocking on theirs to help them with their MTSS implementation. The interesting thing is that they've been doing it for 15 years, I believe, and so they're well versed in it, but they were in finding new intervention strategies or maybe they were kind of getting stuck in their ways and they found that IntelliTier is providing them with new approaches on how to go through with interventions. 

So that was cool to see, but I think generally it's going to be either an innovative superintendent who's trying out new things, willing to do that, or districts that are a little bit tight on resources and don't have the student support teams or people kind of focused on this area. The way that bigger districts that are well funded have as much. And then I think we will do well in big districts as well because we help a lot with collaboration and organization of the data and making sure that everybody has access to it, given that permission, they have the right permissions of course, within our system to be able to work together and be in line in terms of which interventions are being applied to students.

Thembi: Okay. Now you mentioned MTSS, is that the existing system that districts use to track student data?

Jose Moreno: So that's multi-tiered systems of support, and then PBIS and RTI both fall under the umbrella of MTSS. They're all multi-tiered systems of support basically,

Thembi: And you're offering a replacement to that, like a similar product, but a more streamlined version of those.

Jose Moreno: It's more like a guidance system. So what we do is we learn which framework you've applied within your school, so it could be PBIS, RTI, MTSS, and then we tailor the AI to that framework. If there are any policies or that the districts have or the schools have that they'd like to follow, then we also give that to the AI specifically for you, so it's personalized for you as well. Tell it context about the schools or districts. So we have the fully virtual one, so recommendations are specific to fully virtual interventions and then context about the student historical information assessments to get the academic side of it as well. And then putting that all together, we can make personalized recommendations for your school or district for exactly what you're trying to do. So we're not necessarily limited to one framework or the other. We don't even come in an opinionated way where we say, you should be doing this instead of this. It's more like, Hey, what are you doing? Let's focus in the AI on that specific area, and then we could tailor the recommendations for you based off of what you're trying to do.

Thembi: I like that, attaching to what they already use and just making the process more effective and efficient.

Jose Moreno: Yeah, exactly.

Thembi: Okay. So when it comes to AI, there are lots of concerns about equity, bias, accessibility. You said you don't come in with any sort of judgment on which system they're already using. You're just saying, Hey, we want to assess and then use our tools to help you do things better. Do you also sort of come in with those thoughts about equity, bias, and accessibility in mind?

Jose Moreno: Hundred percent. Hundred percent. All of these AI models have been trained on biased data. There's no way around it. That's just the reality of the situation. The good thing is that they've also been trained on how to fight bias and focus on equality. That is part of the data that they've been trained on as well. It's not like there's any malintent to say, don't train it on this information, but do train it on this biased information.

Thembi: Okay...

Jose Moreno: It's just that there's a lot more biased information in the data that's being used to train it. And so it just tends to lean that way a little bit more. My take on this is that a lot of times this bias comes in from teams or higher ups that are just not thinking about it, and a lot of times they're not thinking about it because they never lived it. They don't really understand it. It's never front of mind for them, right? Well, that's not the case for us. I have lived it. It is front of mind for me. 

And so we do have strategies that we put in place to make the AI aware that it could be biased. And we have things that we put in place to say, don't do that. You need to make sure that you're focused on equality. And one of the big things that we did initially, which I think eventually I want to move away from, is that we're not providing it too many protective characteristics. If a student has ADHD, we do provide that information to try to get interventions tailored for that. But in terms of race, ethnicity, all of that stuff, we're trying not to put that in. 

Thembi: So you can just focus on, but okay. Sorry, go ahead. I was going to ask a question, but I think you’re going to…

Jose Moreno: I was going to say…

Thembi: …answer it.

Jose Moreno: But even with that, we do give it the student's name and just with a name alone, the AI can make assumptions and it can make those recommendations. So you still have to do some work to try to focus on equality. Eventually we want to get back to the point where we do provide more of those protective characteristics because you can provide even more personalized recommendations, knowing about the student's background, knowing about the environment that they're usually dealing with, and then focusing on what can work for them in that specific situation. And so we will get back to it, but we're being very careful with taking it step by step because we don't want to make any bias recommendations and put kids in a worse situation than what they already had to deal with.

Thembi: That makes sense because that was my thought of knowing what their race is, knowing what their gender is, knowing all these demographics definitely impacts how they're seen. But you're right, it depends on who's doing the knowing and who's making the decisions. Because my assumption is that you've got this AI tool that makes recommendations, but at the end of the day, there's still a human being that has to decide to follow them and follow through with whatever the next step is. And so that's a really interesting set of questions that I'm really glad that you all are tackling because you got to tackle 'em, you can't ignore it. 

Jose Moreno: A hundred percent.

Thembi: And to try to make this work better for the students and for their families makes just a lot of sense for their trajectory of their whole lives as we know. Right. Okay. So are you in the trenches with your team? Are you doing programming?

Jose Moreno: Hundred percent. Yeah, absolutely in the trenches.

Thembi: Is that your vision for yourself as a founder? And if so, do you want to be in that granular work or do you see yourself pushing towards the leadership role in the strategic vision?

Jose Moreno: Yeah, so this is where I said I was going to mention a few things that I do from the Netflix culture, which was there. One of the principles was a big focus on context and not control. And like I said, they did a lot of work to empower the people that were actually doing the implementations as opposed to just having leaders tell people exactly what to do.

And so I really want to bring that concept and philosophy into it. And so me being in the trenches isn't necessarily me telling our UI engineer, our UX designer, exactly where the boxes are going to go for our solution. It's more of a focus of saying, this is our target audience, this is what we're trying to accomplish this, these are the challenges that they're facing. 

We need to make sure that the user experience stays extremely easy and easy to use and fast so that teachers can put in their data collection and get back to teaching in the classroom and not have to worry about anything else so that student support teams can easily digest the information and then make the decisions that they need to make. And so it's more about getting everybody on the same page to see the vision and to focus on the principles that we are making a top priority. And then having the people that are experts in their field, not a designer. 

And so I definitely don't want to go in there and tell people how to design our product. And so it's more about getting them aligned, see the vision, and then having us all execute together into what the final solution that we're trying to get to would be. I guess it never really ends, so at least the next big milestone for what the product's going to be. 

Thembi: But it sounds like you are innovating a certain kind of leadership. I don't know if you would consider it pure servant leadership, but it does feel like you're there and you have obviously experience and knowledge even if you're not in your specialty areas. And so I feel like wouldn't that they would have more respect. You're not just a figurehead. You're in there with them.

Jose Moreno: And once again, I attribute so much on Netflix's success to this because I think my kind of negative stereotype for a lot of companies is that a lot of times the people that are running the company, whether it be CEO, VPs, or directors, a lot of times it's hard for them to put their egos aside. They're very experienced, they know what solutions work for them historically, and a lot of times they kind of come in and say, this is what I want to build, or this is what I know works and that's what we're going to do. Sometimes you have to be able to just step aside and say, I'm going to put my ego aside. Yes, some things work for me historically, but there might be more innovative solutions that haven't worked or haven't been tried and could do better moving forward. And in order to do that, you just have to understand that you may not know what that solution is. And so you have to step to the side and just give everybody the vision and what you're trying to do and then see what comes out of it and trust your team that they're going to be able to execute with what they know also in their specific areas.

Thembi: That alone is next level leadership. There's so many paradigm, the paradigm of the leader knows best, and no matter what the evidence is that the team is presenting, it's like, no, this is what we should be doing. This is my vision. And to say, Hey, I hired you because you're great at what you do, so I want to listen to what you have to say is really, it makes sense and I feel like it's probably going to contribute to your success, but so many people just don't get it. So you've definitely got a leg up with that. So with your platform, is it possible for students and their parents to gain access to the data that the student, that the teachers and the support teams are gathering?

Jose Moreno: So as of right now, we have not built the parent or student side to the product. So it's mostly focused on the administrator side of the world. That being said, we do follow FERPA regulations, and FERPA does require that if parents ask for access to the data or make modifications of the student data for corrections, that we're able to do that. So we do have that side of the world. So if that information is requested, absolutely we'll provide it following all laws. 

Eventually we want to get to the point where we're providing parent side, a teacher side, a student's side, because there isn't always one side to an event. There's what the teacher saw or what the teacher reported, but the student might have a completely different experience. And so I can see a world where the teacher puts in an incident and then maybe feedback or additional information or context is requested of the student, and then the student provides their side, and then you could take all of that information to make a more unbiased recommendation. Because right now there's definitely more bias towards the teacher's point of view and what they're trying to do. 

Thembi: It's very teacher-centered, but that's where the schools are. So I get it, you're meeting the schools where they are first, but I like that you're thinking about later because when I think about, for instance, employee evaluations as adults and in a lot of environments, you get your evaluation, but you get to respond to it. You get to have comments from yourself added to the file so that there's some perspective of the person being evaluated. And even though we're talking about minors, which makes it tricky, it is interesting to imagine as we move on and with this kind of technology that we're thinking about, well, how is the student's voice being represented in that storytelling? So that's an interesting thing to think about. 

We were talking about how you've created this product for teachers and administrators to gather data about students. So in this process, are you managing to gather data about the educators? Are you learning what the educators need, what the administrators need, and how they engage with technology for your own purposes to make the product better?

Jose Moreno: I really love the way that somebody recently mentioned or provided us with feedback of how we're approaching things. We see ourselves as a partner to our users. I'll start off by saying, okay, I obviously come from the tech side of the world and a lot of my team, a lot or almost all the employees within my team also come from the tech side of the world. We know how to build great tech. What we don't know is what you're going through on a day-to-day basis or what any of our teachers or VP or vice principals or principals are going through on a day-to-day basis. That is your personal experience that we don't have insights into.

And so the first thing that we did was we filled up our Advisor Board with people that have been in education for a long time. So we have a superintendent, we have a retired vice superintendent, and then we have somebody that's on a school board. We have a vice principal. So we're getting a lot of different perspectives from people's day-to-day lives within education. And whenever we're going to build something, it's not us once again coming in an opinionated way. It's more us asking them, what's your biggest problem? Where are the issues? Here's IntelliTier right now. What's the next thing that we can do that'll help your day and make your day-to-day easier? And as we move forward, getting more and more users and getting feedback from them, it's this exact same thing. Well, you're in the weeds. You're officially using in intelliTier. What can we do to make it easier to make it better, to make it more valuable for you?

And then they tell us. And then based off of that, it just becomes a question of prioritization and our vision or I guess our question of how many more people are also going through those exact same problems and that helps with prioritization for what we're going to build next. So we're trying to be humble about it. We're not one of these Silicon Valley companies coming in and saying, you need this. This is going to take your schools and districts to the next level and provide some product that's never been tested in any sort of way. That's not it at all.

Thembi: So you don't come in like, I'm from Netflix, baby! I'm from Microsoft! You'd better listen to I have to say!

Jose Moreno: Yeah exactly, "I know what you need." Absolutely not. It's more like, tell me what your shoes are. Yeah, and then let's see if we can build something to help make your life a little bit easier.

Thembi: Did you decide to enter the EdTech world because you had this space, you'd moved on from Netflix, you were done with that world. You said, okay, I'm striking out on my own. You looked around, you talked about looking at ChatGPT, you talked about looking at some other places that you could go. 

You probably were in high demand. Your phone's ringing off the hook and you look at EdTech and you go, that's what I want to do. What made you steer your world towards EdTech? 

Jose Moreno: Two big parts. One, like you said, I was technically leading the account sharing initiative at Netflix. That was after going from skip credits and working on the home rows on Netflix.

Thembi: How could you? How could you, Jose? How could you do that to us?

Jose Moreno: Yeah, exactly. it was three years of being like, oh man, this is not making our users happy. I was always fighting for the users. I will say that I was trying to figure out a good way to make that solution as unbothersome as possible when people weren't sharing, either way. I think part of it was that we wanted to move back towards, or especially I wanted to move back towards providing positive value and building products that help out our society. And so that was a part of going into education. The second part was more strategic of saying or just kind of understanding that the tech companies that have been in the education space may not be as skilled technically because there's a stereotype that there isn't as much money in EdTech as there are in other spaces, like building AI missiles for war. There's definitely a lot more money on that side of the world.

And so from a strategic standpoint, I believe that our company is more technically capable than almost every other EdTech company that's out there. And so if we can just build better products, even if it's the same exact thing, but better experience, easier to use, more organized, then we can compete in the area and that allows us to have a bigger chance of being successful in whatever we ended up doing. And so the combination of those two things and then finally finding a problem that didn't seem to have any sort of actual new solution on today got me really excited about finding something that I thought we can really jump into. And so those three things kind of came together within a few months and it was like, yeah, this is the space that we want to focus on and that's why we jumped into EdTech.

Thembi: Those are just smart business decisions. You can't even argue with that. That's just smart ways to approach it. How long has your company been in existence?

Jose Moreno: We're a year and a few months now.

Thembi: So very new.

Jose Moreno: We're pretty new. Yeah.

Thembi: Are you seeking to build out really fast? Do you have a handful of employees? Are you grabbing people?

Jose Moreno: I'm not trying to get to the point where we IPO the company or anything like that. The hope right now is to build a sustainable company where we can continue to build at our own pace and provide value, but without having that super high pressure situation of having to grow into every single district and school within the next two years or else the company dies and everybody goes their own way, that's not the focus at all. Let's build something sustainable. We want to grow, but at a pace that's necessary as opposed to a pace that everybody gets greedy and really wants to get as big as possible as quickly as possible. And then the hope is that if we can do that, then we can just chill. We can continue to build at our own pace. There's no need to have any sort of end game. It just becomes our day-to-day lives where we're doing things that make us happy and things that provide a positive impact on society, hopefully. 

Thembi: Is this anti-capitalist capitalism? Is this neo-capitalism? What's happening here?

Jose Moreno: Exactly. I think this is the new generation, right? That's what I'm kind of hoping. This is the next generation of businesses that come in. At least for me, I'm definitely not thinking about it in a super capitalistic way.

Thembi: I love that. So your AI model -- is the decision making process transparent? Isn't it sometimes hard to understand what AI is doing? I know I read something recently where the writer wrote about AI that people were able to see the decision making process and there was apparently some deception there. I don't know if you had heard about that, but this idea of the AI like, oh, don't tell them how I'm figuring this out. I'm going to lie to them. Do you have a transparent AI process and is that something that you are able to explain to educators or is it proprietary and you're like, just take the product.

Jose Moreno: It's not proprietary. I'll say what we do. So a lot of these models that have been built either by OpenAI or Meta, they are models that are now just kind of being released for use in different ways. And so what we have is a base model of what OpenAI has built, but we're hosting it in our own Azure, Microsoft Azure servers. So we stay completely within our own environment, so we don't make any calls over to OpenAI. We don't transfer any of our data over the Internet like that.

We have the model on our side. Then what we can do is we can shape that model to be more specific to the exact use case that we want. And that's what we've done in terms of training it on MTSS and PBIS and RT. So that training process that you were talking about was more for that base model, and I've seen some of the news around that and it is kind of scary. The question that I kind of gotten before is do I think Skynet from Terminator is possible? The answer is yes. I do think it's kind of possible. Oh my gosh. And that's part of the fear, but I don't know. I'm not completely into that world right now, and so we'll see where it goes.

Thembi: You're using your powers for good, so you're--

Jose Moreno: We're trying to leverage these models too for good as opposed to training it to the next day. I think a lot of it becomes a question of right now you're pinging it for answers and it goes into a little bit of a black box where you don't really understand what's going on, and then it spits out answers. That is absolutely true. Nobody fully knows exactly what's going on in there to get back the answers that are coming out. And it's similar to the brain, and at the end of the day, these AI models, they're kind of representations of the brains. They're our own take of what's happening within the neurons within our brains, and you just kind of give it information and then something comes out on the other side and it just seems to work. That's honestly the philosophy behind these AI models that were built. And so they've just gotten so big and so complicated that you don't really see all the math that's going on or how it's able to organize the information that's within there because it's just not comprehensible at this point.

Thembi: Well, no, it makes me wonder though. It makes me wonder if one were to say, okay, well, I want to get around the black box problem, could I just feed the information back into the same AI model and say, okay, now show your work. I would do a student, an old school math class. Okay, I see that you found the solution, but show the full process of your work. Do you think it, well, I guess it's like a circular thing, right? 

Jose Moreno: Yeah.

Thembi: Because then it'll say, yeah, here's the work, but I still didn't see the process.

Jose Moreno: Yeah. Well, we know what ends up building these models, and it's basically a whole bunch of matrix multiplications and math through many, many, many layers of just doing math to try to fire what they're calling neurons and different parts of the model to try to get that to go. The way that it works is the way that the training works is that you give it some information at the beginning and then it goes through its own little math sequence and it spits out an answer. And if the answer is wrong, then you give it the answer and you say, go back and modify all those little multiplications that you did to get to this answer. And so the next time you feed that same information in, then all of a sudden the right answer comes out because it went through and it fixed all the multiplication that's happening internally, and then you do that again, and then you do that again, and then you do it again. And so all of these neurons that are kind of interacting with each other just start to learn how they should be adjusting in order to get the answers that we're telling it are the answers to the problems that we're giving it.

Thembi: So maybe that's the closest we can get.

Jose Moreno: Yeah, exactly. That's how that process works. But like I said, once you start doing this at the order of the entire Internet where you're feeding in the entire Internet into these models and training and all that, now you look at all the math that's inside of these models, and you're like, okay, I don't get why this number is negative 13 and this one's 7.35, but at the end of the day, it's just a result of all of that information that you fed in and then trained it on and what came out of it.

Thembi: Okay. So let's talk a little bit about investments. You talked about not rushing to an IPO soon, and so you're bootstrapping your startup and you're not seeking out big investors and you're not out there. The way that you're building this, it seems like is from a place of really trying to solve these problems and develop a proof of concept that you can then expand and scale out. So why did you choose this approach and do you expect to continue in this direction or maybe are you open to later down the line an IPO, maybe further, further or not your thing?

Jose Moreno: I think eventually if it heads into a direction of IPO, then that's just the direction that it heads. Like I said, I like the concept of being able to settle in and just be like, yeah, I'm happy doing this day to day, but in reality, my personality is like, what's next? What's next? What's next? We already got pinged recently about building a side project to help out students that are in juvenile halls and try to get that system a little bit more organized. And so I found that to be interesting, but the more of these products that we end up building in order to help out different areas, obviously the bigger we're going to get. And then eventually there's some size that you kind of get to, but so we're not focusing on it, but if it naturally heads that way, then we'll just have to have those conversations. I have bootstrapped it for the year and a half. We're off to our sales season right now and we have a decently good start now. So the hope is that we could get it to a point where it's just sustainable in terms of the revenue coming in, paying out for all the costs, people's wages, and if we get to that point, then there's no need to look for funding. We just kind of continue to grow, and that's sustainable fashion. And so that's our focus right now. Hopefully we can make it work.

Thembi: What a beautiful concept. I love that. It doesn't have to be too big.

Jose Moreno: Yeah...

Thembi: It needs to just work and be sustainable. I love that. So you just talked a little bit about other products that you could be creating or working on for the kids in juvenile justice. Are there applications beyond the education world that you think that some of your products and frameworks might address?

Jose Moreno: Yeah, it's possible. Right now it's not our focus. There's definitely a lot of different things that we could do. The way that I kind of see it is that with this generative AI technology almost, I would say like 80 to 90% of the software that currently exists can probably be rewritten to put in or to provide additional value in a way that wasn't possible just three, four years ago. And so you can focus on rebuilding products with additional value using this generative AI technology in almost any space. So the options are kind of unlimited right now, but we're liking what we're focusing on right now in education. And like I said, we're not trying to go for the biggest return on investment right now. We're just trying to provide positive value, and so we'll stick within education or this space of the world while we can help out.

Thembi: You're in your early phases of your company and building it, developing it. I'm sure you're refining your mission as you go. Let's just imagine, I don't know, 5, 10, 15, 20 years down the line and the mission of your company is fully realized. What does that look like for education? What do you envision that world looking?

Jose Moreno: I think the biggest thing is we get rid of your biggest headaches. I hear that one of the next areas that I find interesting and just fun to try to solve is the scheduling problem for kids for the following year, right? There are requirements for classrooms that they have to be a part of. There's limitations around class sizes and teacher resources, and I hear that's quite a pain to do every single year. And I think that that's something that AI can help out with significantly. And so just kind of taking that iterative approach, if we can take the biggest problems out of your day-to-day lives and take off some of that over-reliance on our underpaid teachers and just have them focus on teaching kids, I think that that's really a great place that we can end up in where I can potentially see education being in 10 to 20 years and whether or not we're a part of it, we'll see is really focused on personalized learning. Having kids learn at their own paces where what's teaching them can dynamically change in and out or speed up and slow down depending on how it's interpreting those kids. Being able to process that information and being able to recall it a day a week later, kind of come back, reteach some things, move forward if the kids obviously has it. And then just having kids learn at their own paces. I think that that's a super interesting concept and one that I would love to see where it moves forward.

Thembi: I mean, we can see now that in our systems even we think of ourselves when we were children going through schools as well as any of us with children, seeing it from any angle that the classrooms have too many students in them. You've got students at all different levels even though they might be the same age. We have this time-based approach to education where students learn at different paces, they learn in different ways, they need different support. So I could really see where personalized learning with technology, AI based technology could really, really benefit students and help them move at a pace that works for them so they're not being held back if they can move faster or they're not being propelled forward if they need more time, which I really love something that's really great.

Jose Moreno: And something else that I'm really passionate about as well is this whole English as a second language kind of concept right now. My understanding at least in California is that you have to pass a certain test in order to stay within the same resources that all the other kids that know English have in order to keep their learning situation moving forward. If they don't pass that test, then there's an overfocus on teaching them English and underrepresentation on teaching 'em math and geography and everything else that they should be learning during their time there, which only helps them fall back more and more because they're not getting the same resources as everybody else. My view on this is that language is just a form of communication. English absolutely is important and should be a focus of their curriculum, but if they're not well versed in English, it doesn't mean that they shouldn't move forward into algebra and calculus and all these other history classes and everything else. And so I wonder if there's also a way for personalized learning to be in their specific language while also teaching them English to get them to the point that they need to or that English side of the world, but don't have them fall behind in every single other subject while they're in that process. And I think that that's a pretty interesting concept and something that would help out quite a few kids.

Thembi: That's something to really think about because typically the process is immerse them immediately in an English-speaking environment and force them to learn English. And in a lot of cases, the message that gets sent to young people is your first language doesn't have value. Your first language is something that you should feel negatively about and you need to, English is better and you should focus on that. That's how you're going to make it versus leaving space to celebrate and appreciate the culture that the student comes from, the language that the student speaks and saying, okay, and also English. So I love that you all are thinking about that, and as you continue to explore what's happening in schools, obviously you're going to have data that comes back to you that'll probably feed into your ability to make solutions for this.

Jose Moreno: Exactly. That's exactly it.

Thembi: That's exciting.

Jose Moreno: So we're getting familiar with the space, and then we once again, just figure out how we could potentially help.

Thembi: Okay, so y'all have pool tables and ping pong tables and all that stuff?

Jose Moreno: We don't, no. I am a hundred percent on the Netflix side of the world, like, Hey, come in, do your job, and then go home and go with people that you actually like.

Thembi: That's great. I love that. Straight to the point.

Jose Moreno: Yeah.

Thembi: So the final question I'll ask is about self-care. This is hard work. Being an entrepreneur is not for the faint of heart, and so I know you're probably working a lot of hours. So how do you prioritize self-care so that not only can your company be sustainable, but you can be sustainable?

Jose Moreno: So I'm really pushing myself to exercise a lot more regularly and be disciplined on that side of the world because I think that helps out with stress, being able to get some sleep in making sure that your body's just working the right way. I do sit at my desk almost the entire day, which is really horrible. And there's some studies that show that it might be worse than smoking. Really? Wow. I actually wrote, I did a speech on this at some point a few years ago, and I entered the speech by saying at this point, get up, go outside, grab a cigarette and smoke because it's better than just you sitting there. So making sure that I'm getting out for jogs or walks is extremely important. 

I do want to say that I underestimated the mental challenges that I thought that I was, yeah, that I assumed I was going to see as a part of running a company, especially initially, you have this concept, this philosophy, this vision that you're trying to get other people to buy into, and you're the only one that truly believes in that vision. I was able to bring people on board to the company pretty quickly, so there were other people that shared that belief in the vision, but the market, when you don't have that sort of feedback from that market to say, yeah, you're doing a great job, or they haven't tried out the product and started giving us that positive feedback, basically all you're doing is investing a lot of days into building something that you don't know whether or not it's going to work out. 

You see money going out. It's a very tough place to be from a mental perspective, but I'm an unrealistically optimistic person or irrationally optimistic. That's irrational optimism. And so I always convinced myself like, yes, we're doing good. We're doing good. The feedback that we are getting is positive. We just got to keep pushing forward and get to the point where we can get that market validation, which we're starting to see more now these days, so it becomes a little bit easier. But I definitely underestimated that part of it. And you have to make sure that you're able to get yourself back into a zen mode back into the place where you believe and are willing to continue to push forward. Otherwise, you're just going to give up in the middle of it and arguably a lot of times too early and just say that it failed, when in reality you just might've quit a little early.

Thembi: Wow, that's really important. So it sounds like because you're reflecting on that and thinking about that, that you're putting things in place to keep you going.

Jose Moreno: By the end of last year, usually when I was working at big companies, I was able to take vacations and go traveling, but last year where this was a real focus of mine for the entire year, I was unable to take a vacation. Anytime that I try to step away, something would pop up. And so in one way or another, even if I'm taking a day off, even if it's Saturday or Sunday, I would get some sort of message which brought me back in. And so I did feel like I went almost the entire year without taking any sort of break, and by the end of it I was feeling burnt out. And so for the holiday breaks I made sure, good thing is that all of the schools and everybody's also focused on holiday breaks. And so I thought it was a good time to just shut down the company and just be like, alright, product's working, everything's good, everybody, go home. Enjoy your holiday break, get some rest. That's great. Next year's going to be big year. And so I was able to get to two and a half weeks where I was able to disconnect and that helped significantly in terms of getting refreshed for this year and get the ball rolling.

Thembi: Well, that's great. That's great. I'm very much looking forward to what happens next for you and for Neulight, I think that there's something really inspiring about you leaving that huge world that affects so many millions and millions people and making a conscious decision to make education better, because we know that our education system is very, very stressed, and anything that we can do to make our students' learning experiences better and set them up better for life is just such a win. So thank you for doing that. Thank you for making that choice.

Jose Moreno: Thank you. I'm excited. Yeah, I'm really excited about the entire thing. And if anybody's interested in exploring whether or not intervention recommendations or we can personalize the intervention strategies for students a little bit better, please contact me or check out our website.

Thembi: Yes, indeed.

Jose Moreno: We have a free version on the website. At the website is new light.io, N-E-U-L-I-G-H-T.io. The difference between our free version and our paid version is that we create a specific school just for that one user and that user can't share out a login with anybody else, so nobody else can see the same data. And we also limit it to 10 students. And so it's a good little demo for you to go in, put in some student information, ask the AI for intervention recommendations and see if you like the way that it's personalizing those recommendations. And if you do, then contact us and we can start onboarding the school or district pretty quickly so that you can get entire collaboration, collaboration going for student support. We're pretty easy to contact. I'm happy to chat about anything, especially negative feedback. I really thrive on negative feedback because it just makes me just figure out how we could push forward and make a better solution. So if you have anything that you want to share with me, more than happy to talk about it.

Thembi: Awesome. Thank you for talking to me today. Looking forward to what happens next for Neulight.

Jose Moreno: Thank you for having me.

Thembi: Jose Moreno, everybody, and this is KeyBARD. I am Thembi. Until next time...