Want to join the conversation? Connect with us on Tribe.
Find “Founders in Arms” on Apple Podcasts, Spotify, and YouTube.
Transcript of our conversation with Jay Shah:
00:00:00:00 - 00:00:19:06
We're building screenless AI native phones for kids. So how do you how do you keep kids connected to their loved one safely? How do you keep kids connected to the future? Like AI And how do you deliver all that in a very kid, centric design and method looks like that. Once a hardware founder over as a hardware founder I love hardware.
00:00:19:08 - 00:00:39:08
I was telling Jay about this. Yeah, I'm gonna like. But I feel like. Raj you’re often saying how hard hardware is. It's hard. Yeah, it's often hard. But I also think, like, it's, it's important, you know, like, it's, to deliver software, you need good hardware if you assemble the product. So, like, even if the pieces are built out of country, but you assemble the product in country, like in Canada, let's say, and load the firmware in country.
00:00:39:08 - 00:00:56:09
That and the firmware was written in country even technically that passes the rule. And you and then you can say, you know, made in Canada. Did you manufacture your stuff in, in China, Raj. In person. Yep, yep. What would you have done right now if like you, you were still there, we would have just gone out of business so we could.
00:00:56:11 - 00:01:11:15
There was no way around it. Yeah, there's no way around. Because like all of our tools and stuff, we're in China. It would be very difficult. You know, it takes years to make those tools. Why is why is Google Home and Amazon Alexa like, why did they not achieve greatness? Right? Like, what what a missed opportunity, right?
00:01:15:11 - 00:01:20:07
Hi, everyone. Welcome to the Founders in Arms podcast with me, Immad Akhund co-founder and CEO of Mercury
00:01:20:07 - 00:01:22:18
And I'm Raj Suri Co-Founder of Lima and Tribe.
00:01:22:18 - 00:01:24:06
And today we have Jay
00:01:24:06 - 00:01:26:18
Shah. Well, welcome to the show, Jay.
00:01:26:18 - 00:01:28:06
Jay, you and I go way
00:01:28:06 - 00:01:35:06
back. I know Immad if, you know, but I was, Jay's orientation leader,
00:01:35:12 - 00:01:38:18
And what to do? Oh. That's funny. Yeah. I don't know the.
00:01:38:18 - 00:01:39:18
there were 2006
00:01:39:18 - 00:01:41:06
or 5 or something
00:01:41:06 - 00:01:41:21
like that.
00:01:41:21 - 00:01:48:18
2006. Yeah. I was just a kid looking up to Raj. He was, you know, frosh leader, and he was like. He was like a legend for us.
00:01:50:06 - 00:01:53:06
Still, is still. is.
00:01:53:06 - 00:01:55:18
Jay? I was working on a cool new startup.
00:01:55:18 - 00:01:57:06
Really
00:01:57:06 - 00:02:02:17
interesting one that I wanted to show Immad. Immad Doesn't know about it. So this is the first time that you're meeting Jay
00:02:02:17 - 00:02:09:07
Jay previously did a company called BufferBox that he sold to Google. When did you do the company and when was this sold?
00:02:09:07 - 00:02:22:20
Yeah. So we started it in 2011. Like ten, 11 ish. It was actually part of our undergrad capstone project. So we turned the project into a startup. And we sold it in 2012, so it was a very quick exit, I think 18 months from founding to exit.
00:02:23:01 - 00:02:24:19
How long were you at Google? After the.
00:02:24:19 - 00:02:27:23
three and a half years on a four year vesting.
00:02:28:07 - 00:02:29:19
That,
00:02:29:19 - 00:02:33:07
And now you're doing a new company. Did you do another company in between? You know, this
00:02:33:07 - 00:02:41:07
Didn't this is the second real kick at the can? After Google, I took some time and and ran the University of Waterloo startup incubator.
00:02:41:07 - 00:02:42:19
So I helped a bunch. Yeah.
00:02:42:19 - 00:02:51:09
So velocity, helped a bunch of, yeah. Just emerging students, entrepreneurs, companies. Quite, quite a deep tech focus there, too. Sorry, a little sidebar.
00:02:51:10 - 00:03:06:07
I was talking to one of the founders that that I helped back then, maybe two weeks ago. And he mentioned you because they're in the firm, from a tech space that really kind of drug development. And they just got deep randomly. There's, like, out of nowhere, like, completely deep bank. Their us
00:03:06:07 - 00:03:07:19
can't shut down.
00:03:07:19 - 00:03:09:14
No, no, no, not from if you guys help them.
00:03:09:16 - 00:03:20:19
Yeah. You guys help them. Like you're the reason they were able to get back up because they got a call from Brex one day of like, hey, your payments not going through. And the founders all confused. Cold calls this bank and they're like, oh, yeah, we just mailed you a draft.
00:03:20:19 - 00:03:22:07
Like, good luck.
00:03:22:07 - 00:03:24:19
So you guys got all their stuff back together?
00:03:24:19 - 00:03:27:07
It was great. He's very thankful. Anyway, sorry. That's a cyber.
00:03:27:07 - 00:03:32:14
Yeah. Jay, you were running this incubator at Waterloo. I'm actually very interested to talk about that as well.
00:03:32:14 - 00:03:34:02
You know, I think, you
00:03:34:02 - 00:03:41:02
know, a lot of cool companies coming out of Waterloo these days. Any notable ones that were there while you. While you were running the incubator?
00:03:42:04 - 00:03:53:14
Do I mean a bunch? AD Hoc Microsystems is a really cool one. Building like mEMS tech for eye tracking with tons of applications, I forget, like, the speed at which they can track is but some ridiculous speed. And then they're like.
00:03:55:02 - 00:04:01:11
Exactly, exactly. Yeah. Like really, really high end so that you can, like, focus where you're gonna render, you know, whatever, 100 milliseconds before your eyes even get there.
00:04:01:12 - 00:04:22:14
And to use that tech as well to create like atomic force microscopes at like a thousand times cheaper than the closest proxy in an industry. Like, really cool. So the. Yeah, that's one, ground news was another one that was in our, kind of in that time period, really great news aggregator that's done really cool features around like bias detection, comparison of stories like here's, you know, here's the various ways of presenting stuff,
00:04:22:14 - 00:04:24:02
and amazing, amazing
00:04:24:02 - 00:04:30:13
team ever like science was one of the ones I was just talking about on the, on the banking side, really like pharma, tech, drug delivery.
00:04:30:13 - 00:04:33:02
And so there, I mean, there was a ton,
00:04:33:02 - 00:04:34:14
is like 100
00:04:34:14 - 00:04:36:14
companies a year kind of thing.
00:04:36:14 - 00:04:39:02
And what's your. What's your new idea?
00:04:39:02 - 00:04:40:14
Apparently,
00:04:40:14 - 00:04:46:02
the website will be live by the time we publish this. Hopefully. Maybe you can say the name as well.
00:04:46:02 - 00:04:47:02
It's got the new ideas
00:04:47:02 - 00:04:48:14
smell the new company smell.
00:04:48:14 - 00:04:54:14
that's right. Yeah. So you can visit talk poppy.com. And the company's called Poppy. We're building.
00:04:54:14 - 00:04:57:14
pop pie. Oh, pop. Why?
00:04:57:14 - 00:04:59:02
The latter. Poppy.
00:04:59:02 - 00:05:00:14
Yes.
00:05:00:14 - 00:05:05:14
And so we. Yeah, we're building screen this, AI native phones for kids.
00:05:05:14 - 00:05:07:02
And so, so
00:05:07:02 - 00:05:18:02
how do you how do you keep kids connected to their loved one safely? How do you keep kids connected to the future, like, I, and how do you deliver all that in a very kid, centric design and method?
00:05:19:12 - 00:05:21:14
And it looks like. It
00:05:21:14 - 00:05:26:14
Oh he has a he has a picture of it here. Yeah. You could see it. That's cool. So it's greenness, right?
00:05:26:14 - 00:05:38:02
it is. Screamers. Yeah. So it's a single button. It's all voice. And so, you know, a child could say, hey, call grandma or text grandma that'll record the message. Send that by voice. Grandma can text back, or the kid can say, hey, what's a black hole?
00:05:38:02 - 00:05:47:02
And, you know, the LM will generate a great response that's age appropriate. Parents are in the loop. They get a nightly text of, like, here's what your kid was up to, who's kind of here's who they talk to. Here's what to talk to you about.
00:05:47:02 - 00:05:49:14
So it's a, Yeah,
00:05:49:14 - 00:05:58:16
yeah, but for us. But about the youngest, we try to a little younger and there's just a lot of challenges understanding the kid and, you know, kind of linguistic skills kind of stuff.
00:05:58:18 - 00:06:16:05
And then on the top end, it's basically as as old as, the kid gets a cell phone, right? So that's up to the parent in most cases of when do they succumb to cell phone adoption? And in the fullness of time, I think we'll have, like, multiple versions of this that will that will have a screen like, I don't think your ten year old that's trying to an iPhone needs a fully screened device.
00:06:16:05 - 00:06:17:02
There is a utility there.
00:06:17:02 - 00:06:34:22
Around the age of ten, Apple watches become, like, super useful because you've got G.P.S. tracking. You can text and. But, I mean, I don't know if you're going after that, but the. I guess the additional benefit of that, it's attached to their wrist because, like, kids lose things all the time, but Apple Watches do not work at all with, six year old or something like that.
00:06:34:22 - 00:06:42:14
So I think this is an interesting Dan, space. I mean, maybe eventually you will have a watch version of this. That would be kind of an interesting form factor as well.
00:06:42:16 - 00:06:46:14
Yes, yes. A little harder to make a little. Eric seems to have figured that out on the pebble side.
00:06:47:07 - 00:06:49:14
Is it harder to make I mean, to make
00:06:51:03 - 00:06:57:14
Smaller.
00:06:57:19 - 00:07:02:02
Yeah. How much are you going to sell these things for as a subscription? What's the model?
00:07:02:02 - 00:07:11:02
Yeah. So somewhere in the $150 price point for the device, and then somewhere in the, like, 5 to 10 bucks a month, we're still early days. We're like 20 families using it right now. So that's how early we are.
00:07:11:05 - 00:07:27:02
Yeah, this is a cool. I mean, I'm I would get this instantly. I like this. It's almost a no brainer. I have an eight year old, and, I think she would love to to send me stupid messages. You know, let's go. Do you have to hold down the button? Okay. When you press it, can you let it go for a bit?
00:07:27:02 - 00:07:34:00
Yeah. And so it's push to talk like a walkie talkie. For recording messages. You don't have to hold it down just because I don't like constantly holding down. But. But
00:07:34:20 - 00:07:42:14
I really like the ChatGPT has this, like, back and forth thing that, like, you know, I can just leave with my daughter, and they talk back and forth. It's it's
00:07:42:14 - 00:07:44:02
really nice. I think it'll be a little
00:07:44:02 - 00:07:49:02
annoying to press the button. Anyway. What a UX, feedback without trying it out. Maybe
00:07:49:02 - 00:07:50:02
you make that bit, like,
00:07:50:02 - 00:07:51:14
a little bit more flexible.
00:07:51:19 - 00:08:02:14
Yes. We are. Yeah. I get good feedback. Like we're constantly surprised when we go through the logs and look at usage and listen to the audio, which we have parental permission to do, how noisy the environments are.
00:08:04:02 - 00:08:13:02
detection would be really challenging. Like if we were like without that signal of like the kid is actually trying to talk like often the kids trying to talk and it picks up something, the parent saying in the back and then, you
00:08:13:02 - 00:08:14:02
know, there's
00:08:14:02 - 00:08:15:14
yeah, yeah.
00:08:17:23 - 00:08:21:02
Yeah, yeah yeah. Or
00:08:21:02 - 00:08:26:14
if I, if I try and use it in my car like just the background noise of the car, it's constantly interrupting it. So yeah. Anyway. Yeah.
00:08:29:02 - 00:08:33:14
This particular thing was, is about seven months in,
00:08:33:14 - 00:08:35:02
the startup
00:08:35:02 - 00:08:39:10
we for about a year and a half ago. So we pivoted twice. So this is technically our third idea.
00:08:39:12 - 00:08:55:20
We started an enterprise SAS, so we're a long way from that. But but feeling really jazzed about this, this particular direction. And at the end, the feedback we're getting from the users is awesome. So it's it feels really good. And to be back in hardware like that is my that's my roots. And my co-founder two loves hardware.
00:08:55:20 - 00:08:58:14
So like, it's good to be back here.
00:08:58:16 - 00:09:06:02
Is it not like. I feel like Raj used to be in hardware, and he's like, I'll never touch it again. Is is this a space where you didn't have those, like,
00:09:06:02 - 00:09:07:14
you were like, I
00:09:07:14 - 00:09:09:02
never want to do it again.
00:09:10:02 - 00:09:14:02
No, no. I, once a hardware founder, are over. As a hardware fan, I love hardware, and I was telling
00:09:14:02 - 00:09:17:02
Oh, really? Well, why aren't you doing anything? Another.
00:09:17:02 - 00:09:21:20
No. No, I mean, right now, it just this, the chat stuff I have, I think we have a
00:09:22:22 - 00:09:26:14
Yeah, but I feel like, Roger, you're often saying how hard
00:09:26:14 - 00:09:28:02
hardware is, I guess,
00:09:28:02 - 00:09:37:02
really hard. But I also think, like, it's, it's important, you know, like, it's, to deliver software, you need good hardware. And, in some cases, like for kids, I think this is a perfect application
00:09:37:05 - 00:09:38:14
Yeah.
00:09:38:14 - 00:09:43:00
to software without a screen in most cases. And like, you know, screen time is very controversial.
00:09:43:00 - 00:09:48:14
But this was not going to be controversial. So this is going to be something that parents will probably encourage their kids to use.
00:09:48:14 - 00:09:49:14
I think I think that,
00:09:49:14 - 00:09:51:02
yes, I think that form factor,
00:09:51:02 - 00:10:09:11
too, like the fact that screens are the prevalent technology portal for for kids causes like design choices and business model choices that lead to the screen time metric being the biggest thing for the makers of that software. And then that leads to, you know, the ad model that leads to addiction, and then you lose the purity of what you're trying to deliver for the kid.
00:10:09:11 - 00:10:24:03
So like, I think actually the hardware change here is foundational to being able to deliver like a pure experience for the kid where we're like, we don't want you to use this thing three hours a day, right? We want you to like, send a message of love to grandma, answer your questions, and then get back to being like in your childhood.
00:10:24:03 - 00:10:26:14
Like, not taken out of that, but like screens do.
00:10:26:21 - 00:10:35:02
Was there any lessons from your previous hardware startup that you applied to this, Andre? Like, was there, like, things you were like, oh, definitely not touch this or do more of this.
00:10:36:10 - 00:10:42:14
A no, not exactly no. Like, buffer box was like, like more of, like platform technology, like literally big metal lockers and
00:10:42:14 - 00:10:44:02
train stations and stuff. So very different
00:10:44:02 - 00:10:59:01
than, like, consumer hardware. And I do think like the what your point about hardware is hard. And having gone through it once, knowing that we're just embracing that like like, yes, we're going to have you know, supply chain things and inventory and have to deal with tariffs and shipping stuff, right.
00:10:59:01 - 00:11:18:18
Like even like when I when I came to San Francisco last week and I brought some demo units with me, even just like labeling the devices so they don't get seized at customs because they're not commercial products for resale and they're not UL certified yet. Right? It's like, okay. Like with buffer box, we almost got one of our units stopped at the border, actually, for like demo Day, they almost didn't let it through, so we couldn't bring it down to it to get.
00:11:18:19 - 00:11:19:14
Yeah, yeah.
00:11:19:14 - 00:11:24:02
I guess, you know, manufacturing them in China. Oh, I.
00:11:25:00 - 00:11:28:14
The scam we're at now, you're looking at the manufacturer right now. It's me in my
00:11:28:14 - 00:11:30:02
basement. I made all the devices
00:11:30:02 - 00:11:31:14
so far.
00:11:31:14 - 00:11:34:14
Presumably most of the components are coming from China, though, right?
00:11:34:14 - 00:11:46:14
Yeah. Yeah, like PCB and stuff. But we can, like, nothing that can't be done in China, right? So I think for our next batch, we'll certainly do it in, in North America, we will strive to, there's no there's no there's nothing we're doing that can't be done here.
00:11:46:22 - 00:11:52:14
Actually, funnily, you're importing into Canada, so when you bring it into America, it's actually you don't get hit by the tariffs
00:11:52:14 - 00:11:56:02
Well, we. Yeah, like the the last one
00:11:56:02 - 00:11:57:14
was the technical term. The
00:11:57:14 - 00:12:03:14
last place of substantial transformation is the is like the the jargon for that. So I don't know if we would meet that
00:12:03:14 - 00:12:05:02
with what kind of, you know, the
00:12:05:02 - 00:12:07:02
units we built to date. I don't think we would meet that as a Canadian
00:12:07:02 - 00:12:07:23
origin product.
00:12:08:03 - 00:12:08:14
And
00:12:08:14 - 00:12:13:02
I think as soon as we switch the PCB to, to be made in Canada, then absolutely, we would meet that. Yeah.
00:12:13:02 - 00:12:18:02
Know, that's interesting. Who, who who sets those rules? Like what is enough transformation.
00:12:18:17 - 00:12:20:14
It's literally legally defined like the
00:12:20:14 - 00:12:22:02
country's legally defined. Yeah. So for example,
00:12:22:02 - 00:12:35:16
like I don't quote me on this, but I was talking to another hardware founder who was like, if you assemble the product. So like, even if the speakers are built out of country, but you assemble the product in country, like in Canada, let's say, and load the firmware in country, that and the firmware was written in country.
00:12:35:17 - 00:12:39:14
Then technically that passes the rule and you and then you can say, you know, made in Canada,
00:12:39:14 - 00:12:41:02
even though the PCB was were made
00:12:41:02 - 00:12:42:14
in China or whatever. Right.
00:12:42:14 - 00:12:44:02
Oh, interesting.
00:12:44:02 - 00:12:45:14
Yeah. So,
00:12:45:14 - 00:12:53:02
so actually, like, you would have continued to make this bcba in China, but because of the tariffs, you could not make it in Canada.
00:12:53:20 - 00:13:00:14
My Aaron, my co-founder and I, like we are pretty keen on wanting to make it in North America. Anyways. The like the
00:13:00:14 - 00:13:04:02
but is it influencing your decision a little bit or. Not
00:13:04:02 - 00:13:05:14
really.
00:13:05:14 - 00:13:06:14
Yeah. I
00:13:06:14 - 00:13:08:02
mean that's kind of an
00:13:08:02 - 00:13:14:14
interesting outcome that it like actually ends up helping the Canadian economy there. Yeah. That's.
00:13:14:14 - 00:13:16:14
00:13:16:14 - 00:13:17:14
Yeah.
00:13:19:02 - 00:13:28:18
My wife is working on a fashion design brand, and luckily she's doing the most of the production in India, but she's actually switching out all the packaging from China to somewhere else.
00:13:28:20 - 00:13:43:14
So it's interesting how quickly, like, I mean, some things can be switched over, but for things that can be switched over, like everyone, I mean, you just like with these ridiculous tariffs and uncertainty, like everyone has to get it out of China, basically. Like it's like 245% now. It's crazy.
00:13:44:06 - 00:13:44:14
I think
00:13:44:14 - 00:13:46:02
that's the max for couches and stuff.
00:13:46:02 - 00:13:49:05
Did you did you manufacture your stuff in in China?
00:13:49:06 - 00:13:55:02
Raj in past. What would you have done right now if like you, you were still there and
00:13:55:02 - 00:13:57:02
we would have just gone out of business, so we
00:13:57:02 - 00:13:58:14
couldn't. I mean, Yeah.
00:13:58:14 - 00:14:13:00
Yeah. There's no way around because, like, all of our tools and stuff, we're in China. It would be very difficult to. You know, it takes years to make those tools. Like the the mechanics, the tools to, like, make the unique id industrial design of the of the hardware.
00:14:13:02 - 00:14:28:02
It's very hard to move those tools. People, you know, like, you know, and then you have a very, like, intricate, like, manufacturing line. So once you're studying it, once you've set up in China, it's hard to move. It's not it won't happen easily. So if this everything
00:14:28:02 - 00:14:29:02
But right now. Like
00:14:29:02 - 00:14:30:14
you. Yeah. I mean, but right
00:14:30:14 - 00:14:39:02
now, you don't know how long these tariffs will stay for, right? So what would you do. You just like, pause things and like go into like do some riffs and go into like some sort of like a wait and see mode because
00:14:39:02 - 00:14:45:02
I mean I see yeah.
00:14:45:02 - 00:14:46:14
for it.
00:14:46:14 - 00:14:49:14
And we did that. We were, we were manufacturing in
00:14:49:14 - 00:14:51:02
2018, 2019.
00:14:56:14 - 00:14:59:08
That was 20%. Right. Yeah.
00:14:59:08 - 00:15:19:05
we increased the prices during the contract negotiations I remember those meetings like oh yeah guy sorry Trump for these tariffs on most of our customers were Trump voters. So it was it was interesting to see the reaction to it. Because, you know, restaurant chains are very, pro-Republican in general. And so but so
00:15:19:05 - 00:15:27:02
Sure. But like 245%. It's like basically a blockade rather than rather than a tariff. It's like a who's
00:15:27:02 - 00:15:28:02
I don't know, one would
00:15:28:02 - 00:15:29:14
accept it.
00:15:29:14 - 00:15:35:02
People would just say, I'll wait for Trump to leave office before I buy this product, you know? So like, that's what they would say.
00:15:35:06 - 00:15:47:14
Rajan on the business model for Presto was was there like, when you think about LTV of the customer. And I assume most of that was in the recurring like software. So could you have absorbed that, that 200% like, like and still made it work?
00:15:47:19 - 00:16:09:19
No, no, not in our case. I mean, it, we did it. I mean, you know, we sold it as a package deal. So it's like I say, I think I remember at the time it was like 30 to 40 K per restaurant for a three year contract. And so and the cost was roughly half of that, the hardware cost was roughly half of that.
00:16:09:21 - 00:16:28:11
So the software was roughly half, which is, I think, typical for a hardware software product. Even the iPhone I think is like 50% margins. So, you know, if you double the cost of the hardware, you're, you know, without increasing price or, you can't have a business even if you increase the price of the hardware 50%.
00:16:28:12 - 00:16:48:22
Yeah, you're basically out of luck. So, because, yeah, customers won't absorb probably more than 20% increase. So, yeah, I think, this was going to help many companies out of business who can't move their supply chains. And I think many, I've been reading, you know, on Twitter that many companies have just discontinued products from China.
00:16:49:00 - 00:17:07:02
You know, so and obviously today I was buying clothes for my, my boy, buying clothes like, hoodies on Amazon and I saw many of the hoodies that we normally buy are like, basically sold out and like, they're not you can't buy kids clothes anymore. People are just discontinuing the products. They're just not.
00:17:11:17 - 00:17:18:14
Yeah, yeah. So that's going to happen. But, you know, there's a lot of fallout yet to come. I think we're just at the beginning of
00:17:18:14 - 00:17:20:14
this,
00:17:20:14 - 00:17:22:02
yeah. Yeah. So
00:17:22:02 - 00:17:38:23
You know, Jay, how do you think about, like, I guess people are always saying, oh, hardware startups are bad idea consumer startups a bad idea? You're doing both of them. Like, you know, the do you care about it being a bad idea, or do you feel like kind of your version of it, like overcome some of the objections people might have?
00:17:39:01 - 00:17:42:06
Because I can I can lay out some of the objections if useful.
00:17:42:06 - 00:17:43:18
But, you know, it's normally like,
00:17:43:18 - 00:17:51:18
yeah, the inventory stuff. Hard to scale, hard to acquire customers, easy to get competition, like that kind of stuff.
00:17:52:05 - 00:18:01:06
Yeah. I think all of that is is totally fair. It's also like an empty sandbox because no one's playing in it.
00:18:01:06 - 00:18:02:18
Right? Or, like, relative
00:18:02:18 - 00:18:15:15
to to building software applications. Right. So if, like, if AI's eating software and that's becoming largely commoditized, like where is alpha? Like where do you get unique left. And so I see that as a huge opportunity.
00:18:15:15 - 00:18:34:18
So like, yes, it's harder, but it's also like right for, for like to me it's crazy that that that kid specific tech has not evolved beyond like VTech products or Leapfrog like that really feels archaic technology. When you think about what you and I and people in, you know, have adopted, I have access to every day.
00:18:36:06 - 00:18:38:15
And so like that's like, that's that's the opportunity, right.
00:18:38:15 - 00:19:00:23
Like let's, let's fast forward, let's, let's jump 20 years here and give kids the experience. Like I think anyone probably listening this podcast would agree that kids, you know, in that age bracket right now, let's say, you know, ten or under at this point are going to grow up fundamentally I native. Right. So like are they going to do that on Chromebooks and whatever they're getting in school or there's they're going to be more intentionality about that.
00:19:00:23 - 00:19:15:18
That goes down the stack. Right. And to me, it seems obvious that there should be more intentionality about that. And it should it be some like commodity products with double AA batteries from VTech be to China or something. Right. Like that doesn't feel like the right solution here. So,
00:19:15:18 - 00:19:17:06
yes, I think some of the
00:19:17:06 - 00:19:22:12
structure, but I think there's also this really incredible opportunity and a really starved audience.
00:19:22:12 - 00:19:32:06
Right. Like a really starved user base. Yeah. Now, I mean, some of our investors have expressed many of the same concerns, especially given that invest in us when we were building and process at the beginning.
00:19:33:18 - 00:19:42:18
Yeah, yeah. So it's we've had those conversations, but the more we keep double clicking into this, the more it's like the the love that we're feeling from the users is, is too cute.
00:19:42:19 - 00:19:50:18
Like, if we, if we turn this product down right now, I think I'd have multiple grandparents at my door with pitchforks, like demanding that needs to be turned back on because
00:19:50:18 - 00:19:52:06
the back and forth they get now between
00:19:52:06 - 00:19:59:12
your grandchildren like it's just it's just too valuable to like go away, right? Like, if you know the parent doesn't want to pay, we'd be like, okay, grandma, how much is that?
00:19:59:12 - 00:20:04:06
Namely tax child worth? Like name your price right?
00:20:04:06 - 00:20:11:18
I know if you can get a network effect working, like, maybe the kids can talk to their cousins or friends or whatever, and. Yeah.
00:20:11:18 - 00:20:21:18
100% part part of our early deployments has been exactly that. Like we followed the family network. It's we now have one grandma that's got five grandchildren, like, all talking to her now. And so she went
00:20:21:18 - 00:20:23:06
from, from literally zero.
00:20:23:06 - 00:20:28:13
And now, of course, all those cousins can talk to each other too. And they do. So it's really yeah, it is definitely really cool to see that.
00:20:28:13 - 00:20:41:08
And then like kids in extracurricular activities are now asking like, how do I get one of those? Because I want to talk to my friend as well. So, I definitely think there is an area that very similar to like, you know, the mid 2000 messaging apps like Kik and whatever, where like they just kind of spread like
00:20:41:08 - 00:21:03:06
I think the key with both consumer and hardware is like, you want some sort of network effect. So it spreads without like high cost. And you want the hardware to be like pretty commodity and have most of the intelligence in the software. Like, it sounds like the hardware is like basically Wi-Fi mic and speaker. And I think if the software has intelligence, like it's harder to compete because, you know, there's not that many people that can build great hardware and great software.
00:21:03:09 - 00:21:10:06
Yeah. Exactly. Agreed. Yeah. This is a thin client, so it's there's the magic is not in the hardware. That's just enabling the experience. Yeah.
00:21:10:11 - 00:21:24:04
I think there's going to be a few keys. Right? There's, I think you need to have good design hardware. Like. It's like you should have a personality. Like, I was talking to Jay about this, and, you know, so it should feel like a product that kids want to use and parents want their kids to use.
00:21:24:06 - 00:21:40:06
And then, yeah, you need really good, like, you know, cloud tools and, like, you know, tools on the other end, you know, I think you. Yeah. Jay, you mentioned telephony. Rachel. Like, it connects to, the grandparents, you know, phones seamlessly and in very easy to use interface and,
00:21:40:06 - 00:21:44:06
There's no app like the grandparents sees this as a cell phone number from their perspective.
00:21:44:06 - 00:21:45:18
Oh. So they just to it as a text.
00:21:45:18 - 00:21:50:18
They literally get an, a mass with a voice attachment. They play it, they hear their grandchild's voice, they reply back with the voice message. So, like,
00:21:50:18 - 00:21:52:06
talk to
00:21:52:06 - 00:21:55:15
ought to onboard the grandparent. It's like, oh, your grandchild has a phone number.
00:21:55:15 - 00:21:59:06
Now. Like there's no sign ins. There's no like it's just this just works,
00:21:59:14 - 00:22:00:06
Yeah.
00:22:00:06 - 00:22:01:06
Yeah, yeah.
00:22:01:06 - 00:22:02:06
great idea.
00:22:02:06 - 00:22:03:18
I have
00:22:03:18 - 00:22:16:18
two suggestions for you. Number one, I often find it's hard to, like, get kids like to know what to talk about, and it'll be cool if the thing could suggest things. It's like, you know, like you said, like, ask me about Black Hole, but it'll be more interesting
00:22:16:18 - 00:22:18:06
that there was just a button that says, like,
00:22:18:06 - 00:22:20:07
hey, do you know about, like, black holes?
00:22:20:07 - 00:22:35:18
And like it went into that conversation. So like a I'm feeling lucky tie button would be a cool idea. Another thing, so actually my wife is about to do, us, citizenship test, like the. Yeah, she's about to naturalize, and she just had ChatGPT,
00:22:35:18 - 00:22:37:06
like, ask her
00:22:37:06 - 00:22:48:08
questions over and over. And it'll be interesting if you had, like, a skill tree kind of thing for kids, it's like, oh, you know, this kid is like learning math, and they can, like, do addition, but they can't do multiplication.
00:22:48:08 - 00:23:08:01
Like, let's kind of like, teach them like basic multiplication and like, it kind of had this, like, skill tree and like you as a parent could say, hey, can you can you focus more on this, like using AI as education tool? I think actually, like, would unlock a massive market. Just because like, parents are willing to a spread that and pay for their.
00:23:08:01 - 00:23:17:03
A big yes to both of those. Like, I think as we roll this out, the the platform potential for this where, you know, someone like Class Dojo can say, hey, yeah, I'm going to I'm going to now
00:23:17:03 - 00:23:18:15
deliver some of our content
00:23:18:15 - 00:23:21:19
on here. Right. Like we don't really need to build the math tutoring or whatever.
00:23:21:19 - 00:23:30:03
Like there's lots of people that can do that on the fly. And, and we've actually found kids organically, like, they might come home from school and be like, hey, quiz me on whatever topic they were learning about in
00:23:30:03 - 00:23:31:15
school. Like, we've we've seen that
00:23:31:15 - 00:23:36:08
unprompted by us, which is really cool because the kid wants they either exercise their knowledge or learn more facts.
00:23:36:08 - 00:23:43:15
And they came up with that. That request themselves. The blank slate problem that you described earlier is definitely real. Like on our V1, when we when
00:23:43:15 - 00:23:45:03
we had like the previous version with these
00:23:45:03 - 00:23:52:16
units, that was a big issue for kids, just didn't they was like, you know, you can talk to to you know, a computer got here, like, what are you going to do?
00:23:52:16 - 00:24:01:15
And they just didn't know what to do. So we did give them a press a button and you get an interesting fact thing. It felt like we got feedback that it's just like random. It was two random. It was two out of
00:24:01:15 - 00:24:03:03
context. We need it. There needs to be a lot more
00:24:03:03 - 00:24:07:16
personalization too. But I yeah, for younger kids, we should totally do that as well.
00:24:07:16 - 00:24:08:03
Yeah.
00:24:09:05 - 00:24:23:19
I guess a bit of a fork in the conversation. You know, I sometimes talk to founders that have exited, and, I feel like often they have, like, a little bit of, ego attached to the next startup and, like, they feel like it needs to be really big. It seems like you're very
00:24:23:19 - 00:24:25:07
actually, like, down to earth about
00:24:25:07 - 00:24:25:09
it.
00:24:25:09 - 00:24:40:07
And you've pivoted twice, like. Yeah. How do you think about, like, you know, like when you pivoted and like, you started with this enterprise last thing, like, you know, did you feel like you're failing? Like, how did you deal with that kind of emotional journey as like, you know, second time founder.
00:24:40:21 - 00:24:50:19
Yeah, that is very spot on. That there's, I've definitely felt a lot more pressure. Right. Like your first time around. There was no expectation. Like, nobody.
00:24:50:19 - 00:24:52:07
Whatever. Like. Yeah. Nobody was
00:24:52:07 - 00:24:59:05
expecting you to do anything. This time around, it feels like there's an expectation like, oh, what's what's Jake going to do next rate. And it's probably entirely projection.
00:24:59:05 - 00:25:00:19
Like I don't think anyone actually cares or
00:25:00:19 - 00:25:02:07
is paying attention. Right.
00:25:02:07 - 00:25:15:06
But but it's that it's that self-imposed pressure of like, this time it's got to be better than last time at least, right? And so, yeah, I think that that was that was definitely a challenge. I think we fell prey or I fell prey to some of that in our early brainstorming.
00:25:15:08 - 00:25:29:02
Because because our company came together as more like a combination of the right, the right time, the right people. So like my co-founder Aaron, he was at Google for 17 years. I didn't wander around for a little bit looking for something to do. The timing was right. So so it was the team first. And then we went pursuing that idea.
00:25:29:04 - 00:25:47:07
And I think because of what what you're talking about, I probably over focused on like or what's going to be a thing that can be really big scale. What can what can get to $1 billion? What's going to be, you know, what's in vogue right now, what's going to be fundraising or whatever. Right. And that doesn't mean we ended up with bad ideas, but we ended up with like, bad founder market fit
00:25:47:07 - 00:25:48:19
stuff that like, wasn't us.
00:25:49:00 - 00:26:05:14
Like we didn't feel that passion. So like our previous two pivots, I think those products should exist and somebody should build them. But like, we're probably not the right people to, like, pierce through the world of, you know, that's that's preventing adoption and getting through that, whereas this feels so different. It's like we both have kids, we both want this like we see it every day.
00:26:05:14 - 00:26:33:07
It's like the family life integration, like my, you know, daughter comes home for school and she's off for testing the new feature that we launched yesterday. Like, it just it just works really, really well for all of those reasons. And the hardware element of like, this is us, we are hardware people. So I think it took us or me, probably a year to shed some of that like weight of, of projected burden on what other people were expecting of me and what I thought we needed to achieve and just return back to the roots of like, build something people love.
00:26:33:09 - 00:26:49:07
And just focus on that. And obviously that was like the mantra at work. And like, if I've learned anything at the macro level of startup building in the last 18 months, it's like that is really helpful. Focusing advice, and I shouldn't have drifted too far from it. So I'm glad we are where we are now.
00:26:49:13 - 00:26:52:07
Yeah. Was there,
00:26:52:07 - 00:26:53:19
like, something,
00:26:53:19 - 00:27:06:16
in doing the previous versions, was there something that you, like happened that made you go. Okay, you know, I should really, like, go back to my roots and really care, like, was it, like, how did you end up, getting to this answer?
00:27:06:16 - 00:27:23:07
So the first, the first thing that we were working on, like I said, was enterprise stuff, like when we started going through the early adopter motions and like, having those conversations around, like, soc2 compliance and, like, procure, it was just like, this is so not us. Like, I'm going to I'm going to like, puke after all these meetings with like, just
00:27:23:07 - 00:27:24:07
how
00:27:24:07 - 00:27:25:19
difficult anything
00:27:25:19 - 00:27:26:04
is.
00:27:26:10 - 00:27:37:02
So, like, for me personally, I was like, I can't like this, this is going to work, right? Like it's going to be soul crushing for me. And then there was some other stuff that happened in the market that made that made sense, that we should pivot. The second time around, it was not like it was different.
00:27:37:02 - 00:28:00:03
It was we were building consumer software that, it is a voice agent that you'd upload a bill to. So say like, like a Verizon bill. And then we put our agent would call Verizon on your behalf and negotiate a better rate and say, like, hey, I'm going to I'm going to switch to where, like, and in Canada in particular, because we have like, essentially, you know, a monopoly of, of telcos, there's so much margin to extract from, from telcos.
00:28:00:05 - 00:28:22:04
And so it really scratched our inner rebellion like in the inside of us. And it was it was working in a couple months we save our users like 100 grand. We were taking like 23% cut. But that one was like, okay, this, this isn't gonna scale like that. Like we were in an adversary, an adversary by design relationship where the other party can just choose a different business process and our business is dead.
00:28:22:06 - 00:28:38:11
Like, we can literally can't control that. So, like, I think that was the exposure of, like, okay, here's all the really cool things we can do with voice AI. And then and then when we went back to the whiteboard and said, what is really interesting to us, that voice AI connection into kids, tech and kids exposure and kids adoption was was what kind
00:28:40:01 - 00:28:58:08
And I think one thing that emerged, you know, a lot of the hardware founders actually talk to each other about, like, hardware issues. So, like, Jay is very well connected with Eric, which is going to be on the podcast soon. Scheduling and who worked on Pebble. And like we all talk also about like what the what can go wrong and what, you know, what lessons to learn.
00:28:58:09 - 00:29:16:03
Hardware is always going to be hard, but there's always, I think people who work on hardware always have a passion for, for like, you know, getting it right the next time and, you know, it is an interesting community, actually, of hardware founders. There is, you know, everyone talks to each other and, learn from each other because because everyone it's hard.
00:29:16:04 - 00:29:37:04
It's. And and the space is so big. Like, if you get it right, like you really can benefit the, the the markets are massive, right. And, you know, but you know, I think, Jay, what what are some of the lessons you've learned just from, like, observing hardware over the last decade? You know, you've seen the Pebble sort of, you know, issues, but but there's a lot of other companies do.
00:29:37:16 - 00:29:48:00
Yeah, like I think. I don't think there's a better time to do hardware than now. Right? Like the ability to iterate the the cost of prototype, like the capability you can have literally in your, in your basement to
00:29:48:23 - 00:29:52:16
But what is. What has improved the ability to iterate.
00:29:52:16 - 00:29:56:14
Just even something as simple as how far 3D printers have come. Right.
00:29:56:14 - 00:30:13:04
So I can, I can, I can do like ten enclosure designs in a week and have them in users hands and have them testing like we iterate on this particular button, maybe like 30 times we settle in the right field or architecture, the right force. And it was so fast to do that, waiting for anybody, right? It was like in CAD software here.
00:30:13:07 - 00:30:32:08
Turn around an hour later, I've got the design, turn around, go to a kid, use it like that. That level of rapid iteration, like the hacks accelerator. I think they're still around. But years ago, that was the thing. You go to Shenzhen so you can rapidly iterate because you're right where all the stuff is made. That level of capability is now come into your, you know, basement or workshop or whatever you want to call.
00:30:32:11 - 00:30:55:16
Not that like there's there's limits, obviously. Right. Like I don't have a spare facility in here, but you know, like I started like Volterra, you can actually literally make a PCB board in my basement. Right. Like so these types of things are, are making it much easier to iterate, which is so important for consumer hardware. Right. Like like I mean, it's important for every startup where you're not just navel gazing and theorizing about what's going to work, but actually with users going back and forth.
00:30:55:16 - 00:31:06:16
And so the the barrier of that have become lower makes it so much more approachable to, to do. You do have to wear a lot of hats though, right? Like, I, yeah, it's fun going from being
00:31:06:16 - 00:31:19:10
You named, like, four hats right there. It's like design cad, all the stuff. Okay, so you were saying, iteration is better. Why else was a good time to do a. How do I start a.
00:31:19:10 - 00:31:33:15
I think there's been a ton of stagnation in the hardware world. Right. There's basically, I don't know, like three and a half companies that make hardware. Right. That the entry consumer stuff. And that's lame. Like, you know, they make good hardware. I'm not like, it's fine, I have whatever, I have an iPhone, I have a Google phone.
00:31:33:15 - 00:31:51:04
That's just fine. But like, that doesn't seem like the correct end space for flourishing innovation in this moment of time. Right? Like, shouldn't we have a lot of people with a lot of shots on goal on that trying new types of hardware? And anyway, we are certainly seeing now like rabid and humane and whatever there are, there are attempts.
00:31:51:04 - 00:32:09:18
But that's a relatively recent phenomenon. I think we were in this like hardware dark ages shortly before that. That's like so I think that's kind of it, like there's stagnation. And wherever there's stagnation, there's this ripe for opportunity where you got these megaliths that are slow moving. I've missed on a couple things like, why is why is Google Home an Amazon Alexa?
00:32:09:18 - 00:32:17:04
Like, why did they not achieve greatness? Right? Like, what a what a missed opportunity. Right. As an example of that stagnation.
00:32:17:06 - 00:32:31:16
Yeah. Well, somehow they still don't have basic. I know, like, I didn't understand. They. You can't ask them the black hole question. Like. Like my Google Home is, like, let me give you a Google search result for that. I'm like, are you kidding me?
00:32:31:16 - 00:32:53:21
right. So that's like there's the opportunity, right? Like we could just we could just make better stuff. Right? Even even things like, like the Raspberry Pi ecosystem or, you know, the ESB, like zero hardware, like, there's just, there's just out-of-the-box hardware that you can use now to build stuff, right? Like that. I don't know if that was true the same way the same level, like accessibility five, ten years ago.
00:32:53:21 - 00:32:54:16
Right.
00:32:56:04 - 00:33:10:09
want to talk a little bit about Canada and how, you know, your, Jay, you obviously did. Why? See, we are live in Canada. And, you know, living in Waterloo, too. I mean, I mean, what is one of is still the tech hub of Canada?
00:33:11:00 - 00:33:21:09
Yeah, I would say so. The convergence of talent, I think it's second to none in Canada. And the universe is obviously the big magnet for that. And I would say co-op, as in the university is really the reason for that.
00:33:21:09 - 00:33:22:21
Yeah. There's so
00:33:22:21 - 00:33:25:09
yeah, incredible builders here, incredible entrepreneur appetite.
00:33:26:02 - 00:33:40:21
How are people viewing like, you know, Silicon Valley or just American tech in general? And then also how are people like, you know, adapting to these, these terraces, like the sentiment, changed for like, you know, working with us.
00:33:41:01 - 00:33:46:09
Yeah. It's so interesting. I think the answer you get would be very different depending on who you talk to. Like in the
00:33:46:09 - 00:33:47:21
in the circles of builders that I talked
00:33:49:21 - 00:33:51:09
because I think
00:33:51:09 - 00:33:57:21
we recognize that the the challenges of building in Canada are not because of the last two months,
00:33:57:21 - 00:33:59:09
that they're there are far
00:33:59:09 - 00:34:03:16
more, you know, systemic to, to, to Canada writ large, right.
00:34:03:16 - 00:34:16:21
Like some of the productivity challenges we've had and things like that. So the builders are just like, we got to get our order in house. We got it. We got to fix our own stuff here and build great companies here. And you see this. And you know, when Toby's talking from Shopify in his commentary. Right. It's like that.
00:34:17:00 - 00:34:33:03
That's kind of the vibe that I'm most exposed to in my circles of builders. And certainly, other founders in town. That would be my sense. There are some, of course, directly impacted where like, I'm literally I need to pivot, but to your point of like, I need to pivot my supply chain, otherwise I'm gonna go bankrupt. But it's out of the acute issues.
Share this post