The API of the Physical World: Sy Bohy on the Future of Connected Hardware

Speaker A: Any smart engineer or founder knows there are certain things where it's like what you see at the surface, uh, level is just a tiny fraction of what you're actually getting. And so with enough intuition, there are certain things where you'll never attempt to build this. We're not going to go and implement payments tomorrow morning. Every time I see those 3 or 4% that stripe takes, I'm like, ah, this sucks. But I promise you, it will be a lot more expensive to do this. Our time is better spent on doing the one thing that's very hard. These sort of incompressible compounding advantage.

Speaker B: Welcome to How I Grew this, where we dive into the world of digital marketing and connect with leaders in the space about how they're scaling, evolving, and growing. Hello, everyone. Welcome to How I Grew this. We have an awesome guest today. His name is Cy Boy. He's the founder and CEO of Psych Siem, which is a software company that allows you to control any IoT device using a single API. So we'll dig into what that means together. Adam calls it Stripe for the physical world of connected devices. Speaking of Adam, um, you and Sai go way back. How do you guys know each other?

Speaker C: That's funny, I was thinking about that before. How long has it been? 10 plus years?

Speaker A: Oh, least, yeah.

Speaker C: Really? We share a co founder is how we know each other. Right, Anaya? Uh, we both had the same CTO over different companies. And Sy, your first company was what, two years out of school is that, uh, you founded a company two years out of school?

Speaker A: A year out of school, yeah, yeah, yeah, Some. More or less like a year out. A couple of years out, yeah.

Speaker C: So you went to Stanford, right? Yep, went to Stanford and then graduated. And you're like straight to Nest?

Speaker A: Yeah, so I joined Nest when it was still a pretty small company. So Nest, the thermostat Nest that I know.

Speaker B: My thermostat. Okay.

Speaker A: Yeah, you know, basically, Nest at the time was a really, really special place. Uh, effectively all the people who had brought you the iPhone and then the ipod before that. So kind of like this core Apple crew. They had decided to start Nest on the back of a handful of Insights. One of them is basically when they made the ipod and the iPhone, they basically figured out how to shrink computing components so that they could effectively fit in your pocket. Right. So we went from a desktop laptop and then an ipod and an iPhone.

Speaker B: Silicon Valley compression vibes.

Speaker C: You've heard the story about Steve Jobs dropping an iPhone in his aquarium.

Speaker A: I have not heard about that one?

Speaker C: No, this was a famous. So they brought it to him and he goes, it needs to be small. And they're like, it's as small as it possibly can go. So he took it over to his aquarium, dropped it in the aquarium and bubbles started to come out of it. He goes, there's air in there. It can be smaller.

Speaker B: Oh my gosh, no air allowed.

Speaker A: Yeah, yeah, that sounds like the thing he would probably do. But yeah. So like the crew that started Nest, they shrank the components and then by virtue of shipping a billion plus mobile devices by, I don't know, 2009 or so, the cost of those components had also drastically come down. And the convergence of these two things was, well, computing is going to get embedded into everyday object, anything from your thermostats, your sensors, your cameras, your whatever, because the computing is super cheap and it's also very, very small form factor. And as a result of this, we're going to have effectively computing deployed inside of our homes, our buildings, our cities, the world at large. And now software is going to be able to start interacting with the ph world and coordinate the use of those device and read some of the data coming back and so forth. And I think if you look at that point in time, when I joined them, um, in 2012, they were just getting started, but they basically kind of nailed it. I think hardware smart home devices got completely commoditized. We went from having maybe a couple of connected devices inside of our home in like 2010 to I think now like the average American owns like a dozen or so. It's kind of crazy. And that's not stopping, right? It's just continuing to go like the hardware continues to get more and more cheaper and available. And it's only fitting that like the people that kind of kicked off the mobile revolutions were also the ones that kicked off the smart home revolution.

Speaker B: What was your connection to those people? Did you know some of them ahead of time or was it just like you landed in an interview process with people that you found impressive and latched on?

Speaker A: Yeah, I'll talk real quick about the interview process of Nest because it is very comical. So I had a very good friend who had worked with Matt and Tony at ah, Apple on the ipod team and then the iPhone team. And when I was, as with every graduate coming out of college, you got to start looking at that and so forth and he said, oh, you should go talk to them. I can't really tell you what they're up to. They're very secretive. But you should Go talk to them. And so I went and I interviewed and I was really just going to do an internship over the course of the summer because I wanted to go back and do a PhD at Stanford. And they made me an offer. And at the time I was looking at different things. I was potentially also looking at full time jobs. And the funny thing about Nest is that at the time when I was looking at jobs at Google's and whatever of the world, the offer was a, uh, six figure salary coming out of school. I mean it was kind of like what you got paid as a computer science grad. And then the Nest offer was somewhere like like $39,000 a year, like basically barely minimum wage. But I remember like looking at the offer letter and be like, how did you guys like forget a 1 in front of it? Maybe like a 0. Ideally at the end they were like, no, that's our offer and you're going to take it. And I remember they had a very specific point of view on things. But I had a wonderful professor in school that it was a security computer science professor, but he basically sort of gave this advice of like, hey, when you're going to graduate and come out of this school, this institution, like you're going to get job offers basically thrown in your face with an insane amount of money to go work at these big companies and don't take those jobs. Those jobs will still be available when you're 40 and you have a mortgage and three kids and a car payment. And when you actually need the money, those jobs will still be available. What you should do early on is take as much risk as possible. Calculated risk, well thought through because right now you have a lot of financial freedom. And so for me, I was kind of like putting that Nest offer in perspective. It's like, well, I can take this very nice offer from one of the biggest tech companies or I can go and go work at Nest, get paid. Not very much money, but basically I get to touch the third rail and learn a ton. And the people there were incredible. And so I think in the end the money figures itself out, but the decision to join them was, I think, the right one.

Speaker C: So you come back to iot, which we'll get to. But the natural evolution from starting at a super high tech startup that's inventing IoT devices for the real world is to go off and start your own company in automotive marketplaces. Yeah. So how did you get to Instamotor from working as an engineer for Nest? How do you found a company after that?

Speaker A: I Mean, I've always, like, wanted to start businesses. I would say, like, joining Nest and working there was almost kind of a fluke for me. And being a founder is what I always wanted to do. I think, you know, at the time, our good friend Val and I had seen kind of an opportunity and was like, all, uh, right, cool. Like, we can sort of chase that down. I think what happened in retrospect is like, that company was hard. It wasn't like, super near and dear to my heart. I love working with the crew there. That was amazing. But I think, like, coming out on the other side of it with all the challenges and stuff, I was like, okay, when you start a business, product market fit is important. The other thing that's very important is like, founder problem fit. Meaning, like, this is something I truly want to work on every single day. And I think that this startup that we did four or five years, like, I learned a ton and it was wonderful, but with a little bit more maturity, I think I would have maybe taken a different path. And I think, like, coming back to siem, post Sensor Motor, I took like a year and a half off to just kind of think very carefully about what's important to work on, what kind of gets you to wake up every day and come in and be excited. And I think for me there was still this interest of, like, computing is getting deployed into the real world. The Nest vision sort of happened. Hardware is absolutely everywhere. The one thing that didn't really happen yet is software is not truly coordinating the physical world. Right. A lot of things are still kind of dumb. And that's really because software has a hard time connecting to devices. And so we sort of saw like an opportunity to go and complete the story of Nest of saying, like, well, I think we need to find a way to make devices very easy to interact with for software, for AI agents, et cetera, in the same way that software today can process payments with Stripe, can send sms with Twilio, so on and so forth. And so for me, this is something that's intellectually very interesting to go after. And this is something I'm excited to wake up for every single day. The seam is not going to be, uh, a quick grow it and sell it. I think it's been really intellectually interesting to come in every day. So this is basically how I eventually found my way, the soul of what I wanted to work on.

Speaker C: So Amanda's introduction was my words. But what does SIEM do, in your words?

Speaker A: Yeah, so basically we provide a single API platform to control Just about any device that you can find. So there's the hardware layer. Like, you know, when you look at door locks, thermostats, sensors, serious cameras, this enormous fragmentation of device and brands and so forth. There are thousands of different brands and obviously tens of thousands of different models. I like to think of devices a little bit like galaxies. Like, we discover new ones every single day. And none of these things are standardized, uh, in terms of how you can talk to them, in terms of the APIs, the, uh, protocols. And so it's pretty difficult to integrate. And on top of that, like, manufacturers can be a little bit cagey about granting access to their device for good reasons. Right. Like, there are security implications to some of this.

Speaker C: Right. Secured cameras. Right. You don't want anyone accessing that.

Speaker A: Yeah. Or door locks.

Speaker C: Right.

Speaker A: Like, you want to be careful about who can come in into a physical space. And so what we saw when we started is we had a number of friends that were creating software applications and they were like, hey, like, as part of this workflow that I enable with my SaaS or whatnot, I need to grant access to a house, or I need to help this building save energy and, or adjust the usage of energy, depending on, like, the cost of electricity at a given point in time. And every single one of these people were struggling to get access to the device APIs that may exist. They were struggling to integrate them, they were struggling to make them reliable even after successfully integrating them. And so we kind of came in and we said, well, IoT devices are going to continue to be sort of a growing force. I think right now we're shipping somewhere on the order of like 2 or 3 billion connected devices per year, like in the real world, beyond just smartphone and computers, like just connected hardware, like 2 to 3 billion devices every single year.

Speaker C: Thermostats, door locks, like vacuum cleaners, refrigerators, everything.

Speaker A: And some of them are kind of dumb, right? Like, it's like, do you really need your toothbrush to be connected?

Speaker B: Oh, yeah. But, like, timers for like two minutes is up and it's tracking and you probably get a streak. I've never used one of these, but

Speaker A: yeah, I have, like the oral B. Whatever. And I don't think I've ever turned on the Bluetooth connectivity. But, like, conceptually, like, the thing that's not going to change is, like, the world is going to continue to deploy a bunch of these devices. And, uh, the other thing that's not going to change is the world is going to continue to build more and more software for Very specific use cases. And these two things should be talking to each other. And in Silicon Valley, every time we've seen effectively an important piece of infrastructure that software developers wanted to get access to, banking, payments, telecommunication, et cetera, we've had an API company or a platform company come about and effectively democratize access to that piece of infrastructure. Right. So stripe for payment, plaid for banking, 12 for telecommunication, I mean background checks, the list goes on. And so SIEM is just like the latest chapter or like an iteration of this. It's like enormous volume of devices being deployed, more and more businesses, prosumers, et cetera, adopting them. Every single one of these businesses at this point in time effectively has a software stack of some kind. They're using some off the shelf application whatnot. And so they're going to want to see more and more automation of the physical world, granting access, saving energy and so on. And we're effectively going to be in the middle, kind of like enabling these things to talk to each other.

Speaker B: So when you out and you said, hey, this problem exists, who were you really thinking you were going to be solving that for? Primarily. And I'm curious if that's evolved over time.

Speaker A: Yeah. So I mean for us, like we've always been like maybe a little bit stubborn about we're building the API to the physical world. And so the moment you say API, clearly you're talking to a very technical customer. Yeah, a developer. I think to this day this is still true, that this is still what we're going after. We did have a couple of like iterations where we try to see like, hey, can we go after like non technical buyers? And I think we've had a harder time conveying the same level of excitement for building to that audience versus like kind of building for developers and doing something that's kind of like stripe. Like what's been super interesting is with all the AI coding agents and all the stuff that we've seen the last like couple of years, it has been a huge tailwind for us. I think if you look at the last year we've 5 or 10x the number of API keys that get created every week we see everybody and their mother coding some application. A lot of people are vibe coding. I don't know how successful they'll be, but like it is there. Like everybody is a developer now. It's super exciting because now I sort of see us as being in a nexus of like something very important happening and we're trying to run as fast as we can to make sure that like our platform is complete. It supports every device you need. You can run very complex workflows off of it for whatever application you might be building or agent workflow you might be running.

Speaker C: It's funny, I heard yesterday that GitHub finally hit the 1 billion commit per year threshold and this year they're supposed to hit 5 billion. So exactly like you said, they're increasing the amount of code commits by 5x in one year. You're seeing that same amount. So you're saying is AI is democratizing programmatic access for your own devices. So like, what is an example of what you call them a prosumer. But like, what is an example of what that person is doing?

Speaker A: Yeah, so internally we kind of refer to prosumer. So I think in the world of connected hardware, devices and so forth, they're sort of like devices that are intended for consumers. And then there's devices or systems that are intended for like enterprise use cases. So the most basic example is like, I think you can walk into a Home Depot and you can basically buy a smart lock. And so that would be a consumer device. But if you're running a hotel or a big building or something, you're obviously not going to go to Home Depot. You're going to go talk to what's called an access control. It's going to be a little bit more intense. But prosumers are interesting because we're seeing more and more very small SMBs that are purchasing consumer grade hardware devices but are using them for business use cases. The most basic example is Airbnbs. Airbnbs are, for all intents and purposes, businesses. They grow to a few hundred thousand dollars a year depending on the size of the portfol. And these people are deploying consumer grade devices inside their home, but they're deploying them for the purpose of making their Airbnb operations more efficient.

Speaker C: Right.

Speaker A: Granting access to guests, turning on the heat or the ac, adjusting the temperature of the pool, so on and so forth. And so prosumers for us are kind of like these people that operate businesses of various kinds of sizes, but ultimately they're deploying what are fundamentally consumer grade devices. And that's a very new thing that's been happening in the last like five years or so.

Speaker C: So you'll see someone vibe code their Airbnb dashboard to control their three houses they own in Austin to ramp up the temperature for people that arrive and make sure AC isn't pumping when people leave. So they're doing that themselves.

Speaker A: Yeah, so we definitely didn't See that like a year ago, historically there's a ecosystem of software application like SaaS, vendors that sell software for managing your Airbnb, your hotel, your apartment buildings and so on. But what's really changed in like the last really six months or so is we're obviously seeing these applications still like coding these workflows on behalf of their customers, but we're also seeing like basically operators that are running Airbnbs or apartment buildings effectively coding their own applications to do all the things that you just mentioned.

Speaker C: So that is a small business or medium sized business using AI to basically automate their own internal business problems. Now that's actually very bridge to another thing that we talk about often, but I want to introduce to our viewers and to Amanda, you're doing that for your own company right now, specifically your CRM system. Tell us a little bit about that because I think that's a very interesting parallel.

Speaker A: Yeah, I think we all know about like the SaaS apocalypse. I don't know how overstated it might be, but I, uh, do think the age of personalized software is upon us. I think it's exponentially easier to create the software that matches it is basically fully tailored to your exact needs. And for SIEM specifically, we're really like now running a ground game of creating basically our own internal tools. So two years ago, building a CRM from scratch, that's just crazy. You don't do that. You go buy HubSpot, you go Salesforce if you have to, et cetera. But we saw that with all the Segentic workflows that are now available that we can basically run on our own data and customers and so forth, a lot of the off the shelf CRMs kind of get in the way of enabling these agents to operate across our

Speaker C: operational stack by design or just they're not flexible enough.

Speaker A: They're not flexible enough. I mean HubSpot is a great product so I want to be a little bit careful not to uh.

Speaker B: No slander.

Speaker A: Yeah, well, it's not slander, it's more like it's a very, very flexible product. It's a very impressive what they've been able to do within the constraints of a ui, like the amount of customization and so forth. But I think like for us, like maybe it's because we're super small, but like the CRM is effectively just a database. And then the integrations that you used to need with a CR M to like stripe and things like that, like these are pretty easy to build.

Speaker B: If you sacrifice the ui, then you're able to get a lot more flexible is what you said.

Speaker A: Yeah. And like, I think agentic workflows, like the data that they need needs to be organized, not necessarily completely differently, but like, they benefit from having like a very simple lookup system for like, data information about the customers. There's a lot of things you want to flow back into your data store. And so again, like, HubSpot and others are great products. It's just like, I think for us, when we started working on like, hey, we're going to really run our current sales pipeline through our own agent Tech CRM, leaning back on HubSpot to effectively just use it as a database didn't really make sense for us. And so it was really about like, let's just build things internally. And again, you can do that now. You don't need like a team of 10 developers to go do it. It's like, I'm able to basically push code and make the fixes I need and again, tailor the software and the CRM as we need.

Speaker C: So you built your own CRM system and is it working? Yeah.

Speaker B: What are the pitfalls?

Speaker A: So we basically have a CRM system. We have a couple of like, agents that will run through it and do things with it. It'll do kind of like the standard account executive T, you know, like follow up with customers, check in on their usage. You want to have a pretty clean playbook for this stuff. But Adam would know, like, if you're running a sales team, generally you want your guys to do like very specific tasks, depending on, like, where the customer is in the pipeline. And so agents are actually pretty good at doing this. The pitfalls is like, sometimes the agents won't quite do the right thing. Maybe they'll forget to do something, whatever. And so normally with human beings, you can just walk up to your AE as like, hey, you forgot to do blah, blah, blah, blah. And it's like, I'll remember, like, it's a very easy corrective step. With agents, you kind of have to get into like a debugging mindset. And it's like, why did it do this? And then it's like, you need to look at the chain of thought. You need to do the root cause analysis and then some of the stuff they struggle on. Like, generally it's a function of like, well, we've never really taught it. We never created like documents or instructions to like, do a particular thing. And then I think just generally, like, again, you want to have like, pretty clear boundaries of like, what an agent should do versus not do. Like the agents are like overly eager to answer very deep technical questions that our customers might be asking. And so for that we're like no, no, no. Like we want our solution engineering team to basically answer that. We don't want like a sales, like just like in the real world, we don't want like a salesperson answering stuff that might not be completely correct.

Speaker C: Right, interesting. So when looking at that, you're saying you have a hyper focused product that you're building internally and that will eliminate the need for you to buy, you know, license of HubSpot potentially. Is that fundamentally uncovering a risk for your business? Because if I can program my own app, what about programming around SIEM, like writing my own APIs?

Speaker A: Yeah, it's a very legitimate question and I think what you see in practice. So first of all, getting access to These device manufacturer APIs is non trivial. Again like there's a lot of business development that's involved the ah, manufacturers, IP licensing, legal agreements, et cetera. There's a lot of regulations around Iot as well, especially out of Europe. So you don't just like walk into like, hey, let me go and connect this particular device and be done in like five minutes. It's not that simple. If you do manage to get access, that's really only the beginning of the struggle. I think what comes after is like these devices behave in strange and mysterious ways. Sometimes you tell it to change the temperature and it doesn't and it's like why is it not working? And so you kind of need to have the hardware like you know, we have an entire lab here in the office where it's like we just have to test these things and do stuff. And then once you actually run things in production, these devices again like will fail for random reasons in the real world. And so there's like a whole operational playbook that is effectively encoded in code. And you can't really develop this playbook to make things reliable in like a day. It's, you know, for us it's basically four years of experience and counting. And there's a couple of like good articles and essays on the timelines that AI can compress and then the things that AI cannot compress today, operational knowledge, playbooks, et cetera, these are the types of things where it's like, unfortunately there's no shortcut. Like no AI agent is going to be able to tell you about all the edge cases, the unknown unknowns that are out there. It is for you to discover and in the world of IoT, because you're enabling sort of real physical world operations for businesses. If you make mistakes, those mistakes are very expensive. So that generally it's like, yes, you could potentially do an integration if you could somehow like get the access through the manufacturers and so forth. The practice is like, it is absolutely not worth your time. And I think what's happening for a lot of SaaS today is everybody is feeling enormous competitive pressure to keep up the features and so forth. And so whenever I talk to some of our prospective customers and say, oh, you know, we're considering integrating directly or not, whenever it's a founder, I'll tell them like, hey, founder to founder, don't waste your time on this. You need to go be building higher up the stack, higher up the value chain. If you spend this time building these integrations with IoT devices and so forth, forth while you're doing this, your competitors are going to be building features that uh, basically will put them ahead. And so you want to be like, very cognizant of like where you spend your time.

Speaker C: That's actually really interesting and very similar to what branch does, because Branch, the deep linking and measurement, a lot of people could build that in house. But that's not the hard part. The hard part is all the edge cases, all the fallbacks, all the getting it to 100%, like, what level of

Speaker B: failure for your links are you comfortable with? Or like in terms of measurement, like are you comfortable having X percent in a silo versus like getting that full picture?

Speaker A: I think any smart engineer or like founder knows, like there are certain things where it's like what you see at the surface, uh, level is just a tiny fraction of what you're actually getting. And so with enough intuition, there are certain things where like, you'll never attempt to build this. Like we're not going to go and implement payments tomorrow morning. Like every time I see those like 3 or 4% that like stripe takes, I'm like ah, like, I know, I know, but I promise you it will be a lot more expensive to do this. We use a webhook service, like basically to issue webhooks to our customers whenever, like a, uh, device action happens. Like we send a webhook to the customer server, we pay a lot of money for this and we actually just signed like a large server renewal with our provider and part of us were like, oh, uh, maybe we should just build this internally. And then we're like, no, this is very hard to ensure the SLA and so forth. Like our time is better Spent on doing the one thing that's very hard, these sort of incompressible compounding advantage around sort of building integration as well as

Speaker C: the infrastructure optimization you have on making it cheaper for you to actually operate it. So therefore, when someone's paying and what is the model are people paying per device, how does that work?

Speaker A: I think like every API company with like enterprise customers, we tend to have like a fairly large number of SKUs. But I think the simplest one is like per device per month, like kind of metered usage. We have some folks that are on like per API call type of setup. Just reason for this. Then we also have per unit of specific, per building, per home, per room in a hotel. These tend to be the three setups we have.

Speaker B: Okay, so I'm kind of dying to ask you about something because I know you probably think about the future of connected devices a lot. And something that's been coming up in our world of branch with mobile marketers specifically over the last several years is like, what's that next device that I'm going to have to integrate into my customer journey or start to understand contextually alongside connected TV and mobile? Do you have any thoughts around, like, where you expect that to go? What we've been talking about for years is like, oh, all of a sudden, are, uh, mobile marketers going to have to start thinking about their fridge and how that connects to the rest of the journey? Or is there some sort of insight you've been thinking about where you feel like it's going to be more a part of a connected experience that people have and that marketers then somehow want to measure?

Speaker C: I see what you're saying. Is your car gonna be the next big smartphone thing you can't live without connected or your energy system? Like, I have a Tesla battery system. I love it. It tells me how much I have left, how much solar I'm getting, all

Speaker A: this sort of thing.

Speaker C: Is there a big device? You think there's a revolution coming?

Speaker B: Yeah. What's the next connected device that marketers should be thinking about?

Speaker A: Well, I don't know if there's like one device, but I think from a marketing standpoint, what's going to happen is more like the combination of multiple devices. Right? Like, I mean, basically at this point, like everything connected, your battery pack in your home, your freaking water heaters are connected now, right?

Speaker C: Like, yeah, uh, well, it's smart too. Like if it knows you leave, it's connected to your calendar, it knows you're leaving for a week, it can ramp the water Temperature down and you'll save a bunch of energy costs. Like that makes perfect sense.

Speaker A: Yeah, but I think what's still like, basically lacking is general awareness what's going on inside of a physical space or like the coordination of m. Multiple devices in one go. I'll bring it back to marketing in a second. But let's take a look at hospitality. Let's say you're like a large chain. I think a lot of marketers will basically talk about the guest journey, the experience, the touch points that you want to have with your customers. A very important touch point and experience is when you first enter that room during a reservation. And today it's basically maybe tap a card and you open the door and that's it. And I think what's interesting with the combination of multiple devices within a particular room is could you walk in and uh, obviously the temperature of the room is already at, uh, where you prefer it to be. I like it very cold at night. I sleep better. So it's like maybe it would readjust automatically but not do that for the next guest is the sonos, the music playing something a little bit different that's unique to you. What does the TV say? That sort of stuff. Right. Personalization of the physical space and being able to do it at scale. Not some CES booth that shows you some cool concept, but actually operationalize this in the real world and being able to take your personal profile of preferences from one hotel to the next within at least the context of a chain. And so when you think about this from a marketing standpoint, like, what should marketers think about? I do think in certain industries, especially ones that are a little bit consumer facing or business traveler facing, really nailing the message around the personalized experience. The guest journey, the relaxing experience, I think is pretty key. And I think these devices are basically going to enable some of that stuff to take place. And it's not just gimmicks and like, oh, it's cool, the curtain's open. I actually think I can basically improve the travel experience. And so being able to sort of convey that from a marketing standpoint, I think it's going to be pretty interesting.

Speaker C: Interesting that's so remarkably similar to what we're seeing today is like AI is enabling personalization at scale. And so now it's important for marketers, builders, whatever, to use the data they have in order to be able to make that personalization throughout the journey. So for you, you're like, how do you share that when it's colder, it's better. Well, they look at the thermostat when you're sleeping in the hotel and you have a ramp down to 62. So they know that you like, like it colder. And so when you walk in you're like, oh, this place feels more comfortable because it's colder. And they have that from your previous day.

Speaker A: Yeah, yeah. Uh, we introduced this maybe about half a year ago. So the concept of a user identity, like a person. And again, we're an API for devices, but at the end of the day we're an API for the physical world. And the physical world, it's like, yeah, it's devices, it's physical spaces and it's people. And so the people part, we basically are going to start pushing the envelope around, like, hey, if you help us understand, and I have this customer, John Smith, staying at my hotel and you know, you have this user identity that we're able to track within siem, then we can start attaching preferences. Like, here's the type of music that we should be playing on the Sonos. Here's the temperature. You know, if you have eight, ah, sleep for your mattress. Like, here's probably like what they want for the configuration. Here's the Netflix account that we need to connect to the TV pre arrival and obviously remove it. Yeah, it's a very common problem in hotels. Like people forget to log out of their Netflix accounts. There's like reset setups now. But it's annoying having to log in in the first place.

Speaker B: We have QR codes for that flow because it is such an annoying use case that people are like, we gotta get a hold on this.

Speaker C: Yeah, it's funny how a connected IoT company and a measurement linking company have so much in common for businesses looking ahead. Sai, if you were to give advice to founders that are looking to build and not be eclipsed by this personalization of software at scale, what advice would you give yourself and other founders or builders that are building for the future?

Speaker A: Sure. I think at this point in time everybody should be building. Like, it doesn't matter if you're not technical, you should be building, you should be coding. Right. Obviously I'm the CEO, but like I still push like 50, 60 pull requests a day to the various tools that we have a day.

Speaker C: How much of your time are you actually spending on that, percentage wise?

Speaker A: I love building, so I think for me it's like every waking hours it's like, I don't have a job, I just come into work like to play basically. Like, I'm having a ton of fun

Speaker C: he is, by the way, the hardest working guy I know. Like his working schedule, normal working schedule is like 18 hour days. That's normal. He's bonkers.

Speaker B: But when you do what you love, you never work a day in your life.

Speaker A: Yeah, a hundred percent. It's hard to call it work when like you get this dopamine drip constantly from like the things you can do. But at uh, the end of the day, I think when you think about starting a business. So you should absolutely be building constantly. It doesn't matter if I think if you're a public company CEO or like a uh, founder, like in your own garage, like you should be building constantly. I think Paul Graham had this really great tweet recently that said the only thing that's worse than a CEO that's a little bit too deep in the AI building vibe coding stuff is a CEO who isn't. So you need to be touching the metal, if you will. I think moats, defensibility, what's valuable versus not valuable. Those questions become pretty important early on. You can see the arc of where things are going. And so you need to think carefully about what makes this little company project that you're doing, what makes it defensible and durable. That's very important in the age of AI because basically anybody on the planet can compete with you very quickly. And the most important thing continues to be distribution, marketing and distribution. I think that becomes more and more important. Denver. It's super, super important. I think having, uh, we'll call it some kind of unfair advantage in that lane is probably one of the most important thing at this point. And I don't know if the rules have fundamentally changed from say two or three years ago. That was already one of the most important thing. But now it's like you need to be very, very good at this.

Speaker B: Truly, this is a theme that keeps coming up and I think it's great advice. Last question I'll ask is just what excites you most about the future of connected devices?

Speaker A: Connected devices. I uh, do think we're getting closer and closer to like sort of writing that final chapter of the NEST vision. So again, kind of like computing is going to get deployed at scale throughout the world and software is going to be able to control and coordinate the use of those devices and basically make the world more thoughtful. So again, it's sort of the finesse vision. Like I think part one definitely happened. Part two where like software of any kind, AI, et cetera, that being able to coordinate the physical world and the devices that are in the physical world. I don't think that really has taken place at scale yet. We're getting there. I think that's going to be pretty exciting. And you're going to see entire buildings, entire homes, entire businesses effectively run completely autonomously. It'll be less energy wasted, more, uh, thoughtful experience. A more fair and unjust world. Right. I think what's interesting about the future, just more broadly, is the timeline is about to accelerate a lot. And I think we're going to enter, like, a golden age of science and discovery. And so I think there's a lot of things to go do in those verticals if that's interesting to people. And I think all the progress, at least I saw over my lifetime, we got some pretty cool stuff, like in the last, like, 10, 15 years. Right. We got reusable rockets, electric cars, iPhones. Probably not going to be very, um, much in comparison to what's coming. I think we're going to have some very interesting discovery, discoveries and advancement in the next couple of decades.

Speaker C: I love it. We've got a software CEO who's very bullish and optimistic on AI in the future.

Speaker B: I love the optimism. You're painting a beautiful future. Let's do that. Amazing. Well, I enjoyed getting to know you. Thank you so much for coming on and talking to us today.

Speaker C: Thank you, Sai. Uh, it was great talking.

Speaker A: Awesome. Thank you, guys.

Speaker B: Till next time. Thanks, everyone. Thank you so much for listening. If you like the show, please leave a review wherever you listen to this and share with someone trying to grow their career or their business. Until next time. Keep growing.