Founder Focus
Presented by: Jade Wang, Kevin Elgan
Originally aired on February 10, 2022 @ 8:00 PM - 8:30 PM EST
Founder Focus is a “Humans of New York” style spotlight on the human stories behind diverse startup founders, their life experiences and perspectives, the origin stories of their startups, and the path they took to where they are today.
English
Interviews
Transcript (Beta)
Hello everyone, welcome to another episode of Founder Focus. I'm your host, Jade, and today we are joined by Kevin Elgan, founder of Chatterbox.
Welcome to the show, Kevin.
Thank you so much for having me. Well, and thank you audience for joining us.
If you have any questions for Kevin, see the email address that's down below and you can email them in.
So, you were introduced to us through DigitalOcean's Hash program, is that right?
Yeah, that's right. I think about a year ago, a little less than a year ago, I came across the Hash program and the perks with it and the partnership with Cloudflare, it's been pretty wonderful.
Cool.
Can you very briefly tell us what your startup does? Yeah, Chatterbox is a smart speaker for kids.
It's actually a build-it-yourself, program-it-yourself smart speaker.
So, you can kind of think of it like an Amazon Echo or Alexa, but one that combines cardboard origami with programming in order to teach kids about how technology works.
Nice. Do you have a unit to show us? Yes, yes. I'll show you what it actually looks like.
It's right here. It actually comes flat. It's all laser-cut cardboard and kids fold it kind of like a cardboard origami to build a smart speaker.
It's powered by Raspberry Pi, has a whole bunch of things and then kids actually customize it.
And this is mine. I'm a big Star Wars fan, so a little baby Yoda and had my daughter paint it and that cut out the ears.
How old is your daughter?
She's three and a half, three and a half. And yeah, the whole idea of Chatterbox kind of came from the struggles of being a parent and trying to figure out how technology works and how those two worlds kind of collide.
Can you tell us about the age range of kids who are playing with Chatterbox and programming it?
Sure. So, we launched a year ago on Kickstarter and the people who backed us were really geeky parents.
But when we design Chatterbox, it's designed for kids about nine years old and up.
We want them to have basic math, critical thinking. But really, we have kids as young as six years old who have made skills to teach Chatterbox.
And our youngest user, besides my daughter, is actually a three-year-old child who has a very geeky mom who likes to make skills for their child.
And then we have a whole bunch of very old users.
So, we had a lot of kind of engineers get Chatterbox for their parents, right?
Because it's just much easier to push a button, like every grandparent knows to push a button and say, turn on the lights.
It's very intuitive for them.
So, it's been pretty interesting. Cool. So, while we were getting ready, you shared your screen with us to see what the programming interface can do.
Can you share that with the audience? Yeah, yeah. Let me see. Share screen.
So, you could actually see here that we're using Google Blockly, which is an open source project.
It essentially allows you to have a whole bunch of blocks that do specific things.
So, I mentioned that Chatterbox is a smart speaker that kids program.
Well, unlike Alexa or Google Home, it doesn't know how to do anything.
That's kind of the feature. So, every Chatterbox is unique by default because kids are going to teach it different ways, give it personality or not give it personality.
But the basis is that they connect these blocks, kind of like Lego blocks, digital Lego blocks, to do something.
And we don't really focus on programming.
We focus on communication. Here, you can actually see it says, when I hear, and then there's a keyword, search for, and then a question.
And we do this. Kids like to get answers to all kinds of questions.
How far is the moon? What's the total height of Empire State Building, which you can see over here.
But then we teach them how to use APIs at the same time.
We introduce them to the idea of what is an API.
There's a service out there called Wolfram Alpha. And if you use these integrations, you can create super interactive skills.
So, I'm going to push a button and ask it a question.
How deep is the Pacific Ocean? So, I did this on purpose because when we're trying to get kids to understand what it's like to teach.
So, I asked how deep is the Pacific Ocean when it's actually like search for.
Exactly. So, I can say, search for the Pacific Ocean. So, I messed it up there.
But actually, we'll see what happens. So, now it searched for the Pacific Ocean and comes back with.
And then I can say, search for the distance to the moon.
And we'll see what happens here.
And this is pretty.
And so, in this kind of like little interaction, you can see that, okay, I made a mistake.
And then I went back.
And it's that kind of process of debugging and thinking through, focusing on communication that we're trying to teach kids how to go through.
So, it's, you know, we're trying to teach kids to problem solve. Think outside the chatterbox, actually.
And really solve problems in their everyday life.
Nice. So, aside from Wolfram Alpha, are there, how broad is the external API base that people can access?
So, we have a whole bunch of integrations. We have weather integrations with dark sky.
We have date time so they can schedule events.
We've used this company called Chirp, which is like kind of audio QR codes to handle a whole bunch of like Wi -Fi setup and stuff like that.
And it's kind of magical for kids because, you know, so much of technology is like abstract that, you know, they can't really tell that they're connecting to a device or like how all of this works.
So, you know, we kind of layer audio on top of it. They can play music.
Smart light integration with like TP Link Casa. We integrate with a whole bunch of APIs.
We have podcasts. And then triggers. And then we're going to be coming out with communication.
So, this was one of the big requests from parents.
Oh, like send a text message to my mom or something? Exactly.
Exactly. And so, it doesn't get more real world than that. And, you know, my background is in like ed tech.
And so much of that is kind of confined to a screen and not really tangible things that kids can use.
And that's really, you know, what we try to focus on with Chatterbox is make something real.
And at the same time, you're going to learn something real.
Nice. Yeah. Actually, the discussion about screens is something that I actually wanted to touch on.
So, as a parent myself, the whole screen time debate has been this hotly debated topic across different parenting circles.
And I myself have been on both sides of the debate, right?
Like, you know, on one hand, there's the, they need to interact with people and the real world at this developmental stage.
And then, on the other hand, it's like, you know, work from home parents during a pandemic.
You've got to get things done somehow or like put a meal on the table.
And at least, you know, with PBS kids, they're like learning some letters and numbers and stuff like that.
And then there's the, like, screen time causes obesity, but maybe that's more because of the junk food ads than because of the sedentariness.
And then, you know, like, and then on the flip side, if my parents didn't give me screen time when I was eight, I wouldn't have started programming.
So, I mean, you've been in edtech basically your whole career.
Can you tell us some of the insights that you've gleaned over time about the screen time debate?
Yeah. You know, it's funny. We consulted with this amazing organization called the Campaign for Commercial Free Childhood.
And so, the executive director over there, his name is Josh Golin.
These are amazing guys, an amazing organization that, you know, really start from a simple question.
And, you know, our guiding question is, how do we solve the kids and technology problem?
Because, you know, it's all over the place.
You know, is it obesity? Is it screen time? Is it learning? Is it personal engagement?
And so, we learned a lot from the Campaign for Commercial Free Childhood.
And their guiding question, well, their guiding question is, or not even question, stance is that there shouldn't be marketing to kids.
And so, I love that.
I absolutely love that. For us, we've kind of twisted that a little bit. We believe kids shouldn't know how technology works.
And as technology has advanced, the lines between real and fake have blurred more, to the point where, you know, I think kids should know that marketing is a type of manipulation, and they should be aware of that.
And the more we can kind of engage our kids and get them to understand what is actually happening, what are these tactics, the better suited they are for success in the future.
Same thing goes with technology. If they understand how technology works, they can think about it, they can innovate, you know, around it.
They're not as, they're more in control of their own, you know, decisions.
And so, I think when it comes down to, you know, those questions that you, that we face as parents, which is, I think the basic question is, as a parent, how do you prepare your kids for the future without turning them into screen zombies?
Right? So, there's a philosophy behind that. And I don't think screens are bad.
I really don't. I think what kids do with screens, we should, you know, be a little more aware of.
We choose to, like, for me and my daughter, we choose to minimize screen use.
And, but she still uses screens to, you know, FaceTime with, you know, grandparents or with family members.
And that's totally okay.
I'll never forget that Mattel, a couple years back, came out with this smart speaker for kids, and they marketed it as a nanny.
It's a digital nanny for kids called the Aristotle, which the campaign for commercial free childhood went after them, because I think we could all agree, we don't want a computer just babysitting our kids.
It's like, that's, that's, so it's a very nuanced question. And I think, you know, parents just should kind of, you know, think about, you know, what are their values as parents and apply that to their kids.
But, but yeah, it's a tough one.
It's a tough one. And we try to just be there, teaching kids exactly how technology works, what's happening.
We get into privacy as well. We teach them about privacy.
That's why we have a button, a big yellow button on there, because we want kids to know that, you know what, we're not always listening.
It only listens when you push a button.
So. Actually, so, so let's dig into the privacy thing a little bit more.
So most of, so most, most common smart speakers that people are familiar with, whether it's Alexa or Google Home, they're passively listening all the time, right?
And there was an article recently where there was a pushed, there was an accidentally pushed update to the Google Home that was where it was just constantly listening and uploading.
How, how does the chatterbox differ in, like, where does the processing take place and what's inside the cardboard?
Yeah, yeah, yeah. So inside the cardboard is a Raspberry Pi. I will show you.
So it's a Raspberry Pi. We created a little hat for the Raspberry Pi. There's a button switch and a modular mic, a daughterboard.
And so we combine a whole bunch of different services.
So we still use, you know, cloud speech to text, but we also have on-device speech to text.
Speech to text is how it like understands, it records a voice and then gives you like a transcript of that.
So we still use cloud services, but we focus on privacy.
So instead of the devices connecting directly, we have no user accounts.
So we don't even know who our users are, which let me tell you, makes it difficult to, you know, support them.
Yeah, you don't know.
I guess the only way you know how many users you have is how many units you've sold.
Correct. Yeah, that's, that's pretty much it. And then we have our server, which essentially every request through chatterbox goes through a chat, a centralized chatterbox user.
So our partners don't get that data at all. It just comes from kind of our backend.
And so we do that kind of as privacy through proxy, but that's just for the speech to text.
You know, all of the intent parsing, like trying to figure out like what is, what a user is requesting happens on device.
And for that, you know, we didn't build our stack up from the ground. We were built on an open source project called Mycroft, which is amazing.
I don't know if you're familiar with Mycroft, but they, they focus on kind of privacy, but it's not suited for kids at all.
And so we've done a lot of work to kind of make that more usable and user friendly for kids.
So on that note, are there, I guess you, it's very hard for you to know whether there are sort of privacy advocate type adults who are using chatterbox as the sort of surveillance free version of like a smart, usual smart speaker.
But you didn't mention there are adult users, like people setting them up for grandparents and such before we went live.
You know, that's one of the reasons why we launched on Kickstarter was because we wanted like a initial group of like really like passionate users who tell us the good, the bad, and the ugly.
And we've just been like cranking ever since, but it's really amazing.
The feedback that we've gotten, the fact, you know, it's funny, we were at maker fair last year and I was talking to a whole bunch of kids and saying, Oh, well, do you have an Amazon echo at home?
And they're like, yeah, but you know, Alexis, you know, not very smart, never understands me or knows like what I can do or what I wanted to do.
And then I said, well, you can teach chatterbox. Chatterbox can only do what you want it to do.
And, and the feedback was just amazing. Kids, kids were like, Oh my God, well, I can control lights using Harry Potter, like magic spells, right?
They can make it work the way they want it to work. And that's the important part.
And that's the number one feedback that we get from all of our users is that they get to make technology work the way they want it to work, not the way an iPhone works, the way Apple wants it to work, right?
Or Android phone, the way Google wants it to work.
But, and, and, and it's not like the good old days in my book where, you know, you had that bigger stuff out and you had like tinker and you had to play around and troubleshoot.
It's like, that's, that's kind of, you know, the essence of chatterbox.
Well, yeah, that it's really cool that the, I'm like, wow, I would want to use Harry Potter spells.
And then I thought, wait, are there, are there parents who, or certainly kids have a wide range of accents and different kind of dialects.
Are there, are there packages that can handle text to speech that can account for that kind of, for a wide range of the way that people say things?
Yeah, that's, you know, this was something that we stumbled on because this was one of my first questions was, you know, you hear all this stuff about, about speech to text engines can't, can't understand kids because of their inflections and stuff.
And quick, quick little story. And like early in early days of chatterbox, I was testing this.
I was like, okay, well, I'm going to test it.
You know, how do I stress test? Well, I'm going to test it with an 18 month old, right.
Or if they can like understand her, you know you know, maybe, maybe I'm onto something.
And so my daughter was like in love with the Moana song, you know, the song from the Disney's Moana.
Yeah. And so I gave her a chatterbox and I was like, okay, push a button, wait for the light to come on and say Moana.
And so I just fed that into like a cloud speech to text engine. And she said Moana and the transcription, because I could see the transcription, what it hears came back marijuana.
And I was like, Oh my God. And I'm like, wait a minute.
This is actually great, because it won't do anything else. It will only so I created a skill when I hear Moana, or when I hear marijuana to say, okay, I'm going to play the Moana song for you and then play it from YouTube.
worked 100% of the time.
And it was so much fun. And I got a kick out of it as a parent. My daughter like, was none the wiser.
And she was able to get chatterbox to play her favorite song, something that Alexa wasn't able to do that Google wasn't able to do.
It was like really amazing how by limiting the scope, like building a general system is really hard, but limiting the scope of what it can do.
And then having more tools to adjust its, you know, kind of action that it takes is actually empowering for kids.
Yeah. So we've actually found that, um, you know, it does. You know, depending on like the developmental age of kids and how their speeches, it is a little bit tougher sometimes.
But that's actually a, you know, a great learning opportunity for my for my daughter, daughter, for example, she, you know, it forces her to enunciate.
She likes Mary Poppins, too.
So she's like, play supercalifragilisticexpialidocious.
And she'll mess it up.
And then she goes, and then push a button again, play supercalifragilisticexpialidocious.
And then it'll, that's a hard word. It is, but it's a learning opportunity.
And that's kind of what we're doing. We're taking like these, these initial initially what we thought were going to be weaknesses and turning them into strengths.
So it's a lot of fun. My baby has just said, has recently acquired the words moo and rawr.
But he only sort of enunciates them.
The 2020 has been an interesting year to say the least for basically everyone on this planet.
A little bit. What's 2020 been like for for you are the pandemic stricken parents rushing to keep their kids occupied?
Yeah, so the chatterbox? Yeah, so initially, you know, I've worked with in EdTech for a long time, primarily with schools and education, K-12 schools.
And, and so when I built chatterbox, I love working with schools for multiple reasons.
One, that, you know, I think that's, we're impact focused.
And we really think we can reach the most amount of kids through schools.
And, you know, maybe give them, you know, a little bit of a different experience that makes them think about the world differently.
So, so for us, we had all these pilots, you know, in progress in January and February, we, we shipped everything in December.
So we went live on literally Christmas day.
And, and that was like less than six months after our Kickstarter.
So it was kind of brutal. And then we went live with schools in January and February and March.
And we're just ramping up, ramping up, and then COVID struck.
And then it's just schools disbanded. And it's not, you know, for, for a pandemic that is highly contagious, having a physical device is not really the best thing.
And especially it's, it's hard for school. So we had to pivot really focused towards parents, which, you know, ended up being pretty good for us.
But again, we've been focused for that educational classroom use case.
And so, you know, we've looked at kind of the pandemic and lockdown is kind of a way to just heads down work.
And we essentially launched, we built out kind of like an LMS and like online virtual courses I've done, and I'm doing currently some virtual camps with some of our users just to get feedback, test ideas off of them.
But, and we've also taken what we have.
And this is brand new, we actually haven't even talked about this or announced this yet.
But we've taken what we have, which Chatterbox on device, moved it completely virtual.
So that means a school can get Chatterbox and give all of their students the ability to teach their own Chatterbox in a browser, and then later take those skills and deploy it to a device.
So, you know, it's, it's been challenging, but it's, we've been doing our best to kind of roll with the punches.
And that's all you could really do and hope for. Or Chatterbox on a phone, I imagine that could also.
That's, that's, that's coming. Yeah. So, so, you know, essentially to be able to run our stack anywhere.
And, and that's, we have the foundation for that right now.
So it's pretty exciting. Chatterbox on a Nintendo Switch.
Like that. You know, when you, when you showed me the giant cardboard thing that folds up into the Chatterbox.
I love it. Exactly. I think cardboard is a great, one of the greatest kind of materials ever.
And, you know, it's, and it was really important that for Chatterbox is part like tangible, physical thing, and then part, you know, digital making.
And, and yeah, it's, it's been a great experience so far.
Well, let's, let's pivot a little bit to shine a spotlight on your, your own personal journey up to this point.
So you've been in educational technology your whole career.
Can you tell us about your experiences growing up and how you discovered technology?
Yeah. So I discovered technology essentially through my dad.
My dad was a technology journalist and he still is.
And but in the nineties, he ran a windows magazine, which was like up there with PC mag.
It's like one top two computer magazine. And and really I was exposed to all this like technology, the Palm pilot before it actually ever hit the market.
And, and all of these like experimental computers that actually never made it to market.
And I got to play with them and it was amazing, but I'll never forget my, my first kind of really formative experience was we got this brand new computer.
I think it was like a, it was like a Pentium two or something. And, and I was like, oh, well, what's this little screw.
And then I took off the side and I'm like, huh, what's this?
I'm pulling everything out of this brand new, like fully loaded computer took, took the motherboard out.
I'm looking at the motherboard and then, and then, and then they're like, you know, because what are you doing?
And then I was like, oh, I'll just put it back together. Cause I always loved Legos.
So I'm like, it's like a fancy Legos. And I put it back together and it didn't turn on.
And I'm like, hmm. And I took it apart and then I put it back together and did it turn on.
And then, uh, then finally realized, okay, you know, process elimination, make sure everything's seated correctly.
The Ram wasn't seated all the way in the motherboard and then like figured it out.
And I'm like, wow, I can just figure stuff out and figuring stuff out is fun.
And, and that was kind of one of that experience really stuck to me.
It's like, if you can put in the work, if you really are focused and methodical, you can do anything.
And, you know, years, decades later, uh, here I am, um, as a essentially poli sci major who, you know, laid out a circuit board and, um, you know, figured out design and like how to scale it, manufacturing all of this, all of this is kind of beyond my realm of expertise, but I think of it, my, my core skill as my ability to learn, I can learn anything and I'll figure out how to learn anything.
I love languages, anything it's, you know, I do like 30 day challenges on, uh, learning new skills.
So. Well, kudos to your dad too, for not freaking out. But, you know, I was, I was resourceful and dedicated and I just love solving problems.
It's like, well, there was a previous guest on our show who, um, who talked about, uh, David Cancel from Drift saying that, um, he, it was 10 years into his career in tech before he, uh, met another person who looked like him.
Um, and, uh, I, I saw that you were part of the society of, uh, Hispanic engineers.
Is that right? Can you tell us a little bit about that organization and what that's been like?
Yeah.
You know, it's well, especially this is another reason why I like working with schools.
Um, it's right. I think right now it's a very, you know, the spotlight is on this, you know, with Black Lives Matter and, um, and, you know, trying, um, inclusion and stuff like that.
And, um, I look at schools and, you know, I grew up, um, I was born in LA.
I grew up essentially in New York and in Long Island, it was a predominantly Jewish town.
And I think I was probably the only, you know, Latinx person there.
And, um, and so it was always, you know, I never really, things always seem kind of different.
And then I came here to the Valley and then, uh, the companies I've worked at were all like predominantly white or, um, and it, and working with like, with schools and education, you have the Palo Altos of the world where it's one way, but then you just, you know, you're from the Bay area from Alum Rock in San Jose is, you know, predominantly like, like Latino and, and, you know, it's like students of color.
And, and so that's one of the reasons why I just kind of like focused on education, because I truly believe, you know, if you can teach kids that, um, learn that, you know, through education, you can open all kinds of doors and you don't have to have the schooling.
Um, we have one minute left in the show.
Uh, before we go, is there any, um, pop culture recommendation, whether it's a book or a film that you'd like to share with the audience?
It's a TV show and book, and I highly recommend The Expanse. I just finished the latest episode of The Expanse.
I mean, the latest season, and it was phenomenal.
And I can't wait for the next, um, for the next season. And it prompted me to go check out, um, Caliban's War.
So that's what I'm reading right now. And it is amazing.
So highly recommended. Thank you so much for, uh, for the show and I'll see ya.