Here we are. Another train delay. Not a problem for Claire. Phone in hand, ready to play the free choral rewards grabber. Loose wrist, firm grip. Tax book. Oh, look at this. She's won herself 50 coral coins for free. But now what to spend them on? Free spins or a casino bonus? And... Oh, she's gone with free spins. Lovely choice, Claire. Cash in Coral Coins for free to choose rewards you actually want.
Coral, we're here for it. 18 plus UK. Participate in selected promos to get Coral coins. Coins expire. T's and C's apply. Take time to think. Welcome to the Mr. Beacon podcast. This week, we have an amazing, unexpected conversation where we explore the questions, what's it like to work for Elon Musk? How do his companies achieve so much? with so few people in such a short amount of time. And how do you get Bluetooth to work inside someone's body as part of the neural link system?
This is actually not the main subject of the podcast. It's kind of a gem that I fell upon in my conversation with Kevin DeWalt, who's the founder and principal at Simple BLE. And I agreed to interview him, was approached. by him and his team. And what he's working on is a really cool project for anyone that's building Bluetooth low energy systems, simple BLE.
offers a great way of getting things done faster. And it leverages the experience that Kevin got working as part of the Neuralink team, fusing the... sensing and motor functions from human beings with Bluetooth. So I love this. You go into a conversation, you expect one thing and you get something. completely different. And it's one of the reasons why I love doing this podcast. So this is kind of a bit of an Easter egg in the interview. The first part of our interview does focus on simple BLE.
which if you're a geek and interested in ambient Internet of Things and how to connect the physical world. to AI and the cloud. That'll be of interest. If you are fixated and interested in how Elon Musk runs his companies and how Neuralink works and how they get so much done, then you'll need to... Fast forward into the second part of the podcast to get that. In the meantime, I'm enjoying my life post-Williot. Williot was a wonderful company, pushing back the boundaries of...
battery-free Bluetooth technology. And I left it after almost eight years to start up Ambi. And things are going well. Tons of people engaging us about. this AI app and network. We're fundraising. If any of you have contacts in the VC community that you think would be interested in fusing the real world with LLMs and making AI omniscient. understanding everything everywhere all at once, then point me at them or point them at me. That's it. I'll let you listen to the conversation with Kevin.
It's a great one, including his music selections, his three songs that have special meaning to him. So do stay till the end and thanks for listening. The Mr. Beacon Ambient IoT podcast is sponsored by Williot, bringing intelligence to every single thing. So, Kevin, welcome to the podcast. It's great to have you on. Thank you for having me. I noted that your marketing budget is, well, what was it? A sandwich and three bucks? Yeah, three bucks, yeah.
Very good. Well, I'm pleased to be part of that. Let's do our bit and ask you to tell us what is Simple Beale? What's the company that you're heading up? Yeah, so this is a project that started almost five years ago at this point. To give a bit of a background, I've worked with Bluetooth since I was a college student because it was a, I mean, it's a great protocol. You can do great things with it.
And unfortunately, what happened was that every time I wanted to develop something with it, I was running into some limitations around the availability of good software libraries to use Bluetooth, in particular when it came to... desktop environments there were no good solutions around it and most of the ones i could find online were either abandoned or in the process of being abandoned
or wouldn't use the technologies I needed them to use. I mean, most of my programming is C, C++, and there were no good libraries that were also cross-platform and would be helpful, but there was basically none of that. And after dealing with these problems for so long, I decided, okay, you know what? I'm doing it myself. I know what I want. I know what I need. And I'm going to start working with that.
I think so much of what entrepreneurs do is that it's like frustration. The world should be this way. It's not. No one else is going to fix it. I guess it's going to be me. I'm going to do it. I mean, yes. And that's how it started. I mean, I... It's not a project that was born, I would say, out of love. It was born out of spite. It can't be that something this simple is so complicated. It just can't. And it drove me crazy.
Yeah, it took a lot of effort. The initial implementation was Linux only. I had done some past attempts as part of one of my previous jobs, but we never... the the whole like vision was born uh after a bit after that and the the goal here was mainly to think okay what are the reasonable use cases you would have for bluetooth
Basically scanning for devices, connecting, interacting with services and characteristics. And having a simple way of doing it. Because not having to worry about all the abstractions of the operating system. Every OS implements it differently. Not only at the sort of API layer. But also the architectural layer is a little bit different. So for example Linux has this concept of you have an adapter.
and a peripheral and the adapter connects to the peripheral on macOS for example you have the adapter you scan you find peripherals but then you tell the adapter
Hey, connect to this peripheral. And the notification that the peripheral successfully connected comes from the adapter as well, not from the peripheral. And Windows as well has its own like... mess around like winrt and every os has its own layer of abstractions around that and the way they implement is like completely different and it didn't seem and for me it was uh it was like okay
There's no reason for this to be so hard. There has to be an easier way to deal with this. And in the end, it was just about, okay, let's design the API first. Let's think about what... what i need as a user and then handle all the painful abstractions in the back end and to make it completely transparent and again this was a project that i was
doing for uh for myself to help me at work and it's it was mostly about like i was solving my own need but interestingly i wasn't the only one so more people started like like you started using it getting involved and it became a i don't want to say a big thing because it's not a big thing but it became a thing i mean if you're looking for like cross-platform libraries especially for desktop we're the only ones
It's actively maintained. Right now we have Linux, Windows, macOS, iOS. We're finishing Android support. We have bindings for C++, C, Python. We're adding bindings for JVM. And we're thinking about expanding this to more languages and environments. And it's only because we thought about this very thoroughly early on.
that we design a very simple API where everything else can be built upon easily. One thing I didn't notice that was part of my frustration here is usually what I've noticed is that engineers, when they're building products, they usually push complexity to the user. Say, oh, I don't want to have to deal with this or think about that. So I'm just going to give you all the options and you figure it out. And in reality, that I think is a...
It's actually a bad thing because turns out 80-90% of use cases would work pretty well with sensible defaults. You just need to think what those are. which is not a trivial task, and you find some limitations. But if you have that, then yeah, sure, you can do whatever you want. And you have all the, you can like, once you have a very simple thing that everybody can understand.
and use adoption becomes easy and uh why would you use something else i mean we've had this discussion for example there's another python library that does i mean it's it's pretty good but it's built around like fully async interfaces. So every time you, if you need to use it, you're dealing with async IO, runtime loops and all of that. And there's some complaint.
I mean, that imposes some additional complexity, especially if you want to run a very simple program. And also you're responsible for handling the run loop. It's not handled by the library itself. So you have to think about all those things and you can run yourself into a situation where you go into a deadlock. All that is, again, just because the design is limited in that way.
and again just by thinking how what's a what's a good and again simple way to do this and to how how do people derive value from this building that and then dealing yourself with a complications around it in which yes sometimes if you look at the insides of simple billy things are very complicated because again we have a very we have a very strong contract that we that we keep
And sure, the OS has to fit into the contract. And sometimes the abstractions on both ends don't fit very well. So you have to really think, hey, what's the best way to do this in a way that's clean, doesn't crash, etc., etc. So it seems like one of the things that you bring is uniformity, consistency across different operating systems, which I imagine just makes things simple for people to understand. But what are the...
What are the potential users or customers for this that when you see them, you're like, you'd be crazy not to be using... our product? And what are the ones where, you know, maybe it's an edge case, maybe we can help a bit, but it's less, how would you just define the sweet spot? So I think the sweet spot comes when you as a company are building an SDK for your product.
I think because right now, most companies that build Bluetooth products are only focused on mobile. So they have to maintain an Android version of their SDK. and an iOS version of their SDK. And you can tell by using both apps, they're different code bases. And sometimes features that are available in one are not in the other. And you have to make this split because there's no solution around that. And if you just...
integrate the communication part into your library and then handle everything else and then you build your business logic on top, it becomes a lot easier. So I would say that's the primary use case. It's like, hey... You can sort of generalize that logic into it, which is what most of our customers generally do. They have some application and they want to wrap it and then use it across multiple platforms.
But also this now opens up the possibility of using your device on a computer. So you can now have a desktop app offering these things. You can have like more complicated versions. And there's an interesting example here at Nordic Semi. I mean, they are the makers of this, the NRF 52, 53, and now 54. It's a very popular Bluetooth chip. All their, for example, their DFU libraries, they're built for...
DFU is what? Sorry, firmware upgrade. They're firmware upgrade libraries. They are built separately as an Android and an iOS package. But if you use their client, their app on desktop...
It doesn't work. You have to get a specially flashed adapter that they have to run it if you want to use it locally instead of using your computer hardware. And this is one of the products that we're having the sort of very early... early stage to like okay package this and just hey here's your device you connect to it you can upgrade it and it just works and again it just built around this idea like why is the desktop environment so neglected and i think it is because
it's hard to do these kinds of... Again, it's already hard to do this on mobile platforms, but you're sort of forced to because that's your main contact point with your customers. But now you can also add desktop platforms. You can think of... new use cases more powerful use cases because now you have a device that is not power constrained can maintain connections for longer and also because of that if you want to automate stuff for example say you're building a medical device you can now have
at a test bench where a computer automatically runs your tests with your device without any sort of human intervention you don't need to have an app emulator or to simulate like someone like clicking clicking clicking clicking like it You can just make it work and you can do it at a scale that otherwise is not possible. No, I'm interested in who's using it. Any specific or abstracted to protect the innocent examples of people using your...
I mean, there are quite a few medical devices, medical device companies that are using it. That was the thing that also shocked me the most. By the time we reached year four and sort of guessing from the comms and people reaching out, I was like, oh, crap, there's a lot of companies using this for medical devices. That was something unexpected, but also we've seen cases in entertainment, security, like devices, for example, for scanning passports, stuff like that.
And the list just goes on automotive as well. There are a few names. I mean, on the website, you can sort of see a few of our biggest. No, I saw Fender. attracted my idea so so what does how do you use it's with a guitar or an amplifier or uh yes it's for one of their amplifiers i don't know the exact model but but yeah they are um
They're very, I mean, they're actually, they've been very, very happy with the results because I like the way they present about it. It's like they're musicians first, engineers, what they care about is... building like great musical products that's what they do that's their focus you when you talk to them you can see the spark in their eyes when they're like showing hey we're doing this and that and for them bluetooth is just a thing they need solved they don't need to spend
months and months of time maintaining multiple apps and all that so hey it's just there and the fact that we can help them is very valuable for them and they've been very grateful and also have provided great feedback on how how we can keep improving things so that they get more value out of it. Very good. So tell me a bit more about... the next level down about the facilities that you offer. And you use a buzzword that is very familiar to BLE experts, GATT. What's GATT?
So GAT is the general attribute table. Basically, the way Bluetooth works is when you connect to a device, you present a series of services and characteristics which you could consider.
talking from a sort of an internet-based viewpoint, it's like, okay, these are like endpoints or URLs where you can like interact and send data back and forth. So for example, if you have a heart rate monitor, there's going to be a... there's a service for heart rate monitor features and one of those characteristics within that service is going to be heart rate
so when you read that characteristic the value you're going to get is the heart rate which for standard services like for example heart rate the units and everything is already defined as part of the standard you just connect to a device you know what it is you know how to read it and you can just parse it that's how for example when you when you have one of these devices and you connect it you connect to it with your phone your apple health
can automatically like grab the data store it and show it to you later on because again all this is already standardized and it's uh in my opinion it's a great way of like building devices because you can isolate functionality for every specific endpoint of course there's some limitations around how much data you can send how you use if you have large blobs of data how you can transmit them over but in the end it's uh for i would say 90
99% of all applications that you can think of that require some sort of mobile connectivity as this one is more than enough. Very good. And what is central mode and what is peripheral mode? okay so central mode is when that when the device you're operating initiates the connection to another device you're the on a sort of low level what happens
What's happening is the device that actually sets the timing for communications is the central one. So basically BLE works by communicating over like pre-specified windows of time. So like, okay, we're going to talk here.
we're gonna talk again in say 15 milliseconds at this other like frequency here and then sort of move back and forth and all the timing for this is controlled by the central so it's usually the device that initiates a connection and interacts with the peripheral uh peripheral mode
is the other side of this uh of this engagement you're a device that accepts connections and usually you're providing some sort of service to the central device for example you're a heart rate monitor blood pressure or whatever other thing you can think of Interestingly, once you're connected, the central can also offer services and characteristics to the peripheral.
For example, one of the most common ones is the name. So for example, when you've paired with a device and the device tell you or you're paid with Kevin's iPhone, Kevin's iPhone was actually a property that... the phone present as a central presented to the peripheral and the peripheral actually stored when bonding was happening. So if I've got like a heart rate monitor that I've strapped to my chest to measure when I'm going for a run, that would be the peripheral? Yes.
Okay, and then the phone would be the central. Yes. The main difference in functionality between central and peripheral is when you're acting as a central, you expect services and characteristics to be provided to you. while as a peripheral you are the one presenting these to the world so you need to have a way of that's why functionally it's different for simply really one mode and the other because for peripheral you need to say oh these are the features i provide
And you're responsible for sending the data for those services. So a Bluetooth beacon, got to get back to that, as that was the genesis of this podcast. That's an example of a peripheral. Correct, yes. A beacon is a case of a peripheral device where you usually don't connect to it. So you only rely on the advertising process to exchange information. There are some, I mean, you have both, you have like advertising and scanning, advertising being just the device saying, hey.
and this and here's some data scanning is the central seeing that advertisement saying hey i think you have more information can you send me that please and you have like sort of a second advertisement packet where you can send more data over but um Yeah, but in most cases when it comes to beacons, you deal with non-connectable devices. Yeah, which is why they can scale and they're so good for public spaces and so forth where you're like shopping malls and airports.
I noticed that you offer consulting as well as license the software. Can you just tell us a little bit about how you decided to package and price what you offer? And I'm interested in... How many of your customers end up using your consulting services and how important is that? It's obviously a good source of revenue because there's always a shortage of talent, but I'm interested as an entrepreneur.
Would you say it was important to offer consulting in order to get adoption of your product? Not really. So there's a bit of history here. Simple BLE started as a fully open source project. Basically, once I left Neuralink and switched jobs to another company that was not focused on embedded devices, mainly I'm now working on AI, as everybody else.
sort of my focus on working on Simple BLE sort of faded away because it felt less important, it was less relevant to my work, so I wasn't spending that much time with it. i noticed like very quickly as i stopped being as involved like quality started to decay like features that for example android support was something that i've always wanted to do but now i didn't have the time or now they have the interest in doing it and
A friend basically suggested, hey, usually if you get paid to do something, you'll feel more motivated to do it. And his idea was, why don't you switch the license to something that is like... more restricted for commercial use basically requiring to purchase the license instead of just using it for free and you see what happens if nobody buys it there's no basically the the trend was going to still be the same
But if there's interest, okay, you know there's something, someone cares about it, and it makes sense to invest more resources into it. And it's been a bit of a bumpy ride. I'm not going to become a millionaire by selling a Bluetooth library, but it's a... it's been a very i would say the process on its own has been a lot more interesting in like learning okay what is necessary to grab a small project like this one and actually put it out commercially the way i describe it is
simple ble is not like a big software corporation it's more like your uh i don't know your corner bakery like very small shop we have we do a few things everything's like very It's not industrialized. It's a handmade, organic, natural, if you want to sort of describe it. I mean, it's a small project. I've been able, thanks to the commercial offering, to bring some people, at least part-time.
into the into the ecosystem and we've been able thanks to that this past year we've been able to dedicate almost another thousand hours of work into the project and some additional things that we're working on like hopefully in the future might pan out as like bigger features and this was all thanks to all these companies that believe in us trusted us and wanted us to like actually help them solve the problem with bluetooth because it's not a it's not a done deal and we have
And it's even in cases where other libraries exist, we've had clients just asking us, hey, I'd rather use yours because it's better, it's simpler, and I have someone to talk to that will actually respond if I need help. The consulting part was basically just the thing we threw on top. It's like, hey, okay, we're selling the licenses, but at the same time, someone might need extra help. Let's offer that.
We basically, we divided it into two tiers. We've got what's called a simple VLE consulting tier, which is, hey, I'd like to use simple VLE for my application and I need help integrating it or I need a...
I need a feature for simple VLE. That's how peripheral support was actually born. It's like, hey, I need this feature. You seem like the best one to do it. So I want to hire you. And the advanced VLE is more for... things that are BLE related but are more these are more focused on like things that don't necessarily involve simple BLE itself where it's more about like okay research or hey I have
I'd like to test some integrations with this product. And I want to know, I want you to reverse engineer the API. Things like that. It was more about... it's for basically the rest where we're not as we need basically more more resources to focus on but usually most of the consulting we've sold has been around simple bly itself revenue wise it's been more or less half and half
So consulting has been a big part of it. But at the same time, it's also the thing that has been the most work intensive. So ideally, I'd like to focus more on licensing because it's... It's something more stable. And as we've evolved, we've also improved our license for this process to make it easier for continued support and all that. And we think it's a... And we think also in terms of pricing, we sort of eyeballed it a little bit around, okay, what are other offerings? What's the...
What's your cost on doing it yourself? And again, it's not about becoming millionaires with this. It's mostly about the experience and to basically make sure that the costs are covered and the library can continue growing.
So we try to keep it as fair as possible. We have a startup tier that is... designed to be free it's like sure you sign the license but it's free the idea is for you to get just to get used to it to get involved we also get to know you to understand your use case so that we can build better features for you and then as the company grows we have like two different
tiers that sort of follow that sort of growth. Pretty cool. Let's talk about something that might be keeping you up at night, cybersecurity. According to Vanta's latest state of trust report, it's the number one concern for UK businesses. That's where Vanta comes in. Whether you're a startup, growing fast or already established, Vanta can help you get ISO 27001 certified and more without the headaches.
Invanta allows your company to centralize security workflows, complete questionnaires up to five times faster, and proactively manage vendor risk to help your team not only get compliant, but stay compliant. Stop stressing over cybersecurity and start focusing on growing your business. For a limited time, our audience gets $1,000 off Vanta at vanta.com slash go. That's vanta.com slash go for $1,000 off.
Because when it comes to your business, it's not just about keeping the lights on. It's about keeping everything secure. So... You've experienced firsthand the evolution of the open source approach.
Anything that you would like to say or any observations on where open source is headed? It's so central to what we do, but I assume it's not static in terms of the norms. It's definitely not static. I mean, one of the reasons... trends is like more projects going down the commercial route and i think it's actually a very positive thing and probably a lot of people are going to be very angry at this point But the reality is that for especially niche projects like SimpleBLE,
there isn't enough interest from large companies let's say microsoft facebook google etc to sponsor them because their bottom line is not affected by this so they have no reason whatsoever to put money aside from maybe the generosity of an employee that's willing to be your champion inside the company. And the reality is that because of that, lots and lots of projects of this size end up becoming abandonware on GitHub. And we did a sort of very preliminary study.
And I think around 60% of projects of the sort of same type have been abandoned for more than a year. And you look at the code and sometimes it's very good, like great ideas, very interesting, but you just need... a bit more to like get it to the point where it can can reach that and we haven't seen any and it's not basically i would say from all the open source ecosystem there's that specific niche that is currently very underserved
And I've been talking with people in the industries, even open source industry like Canonical, and they acknowledge that, yes, there's a small portion of projects that are... important they add value but they're not bit enough big enough for anybody to actually care about them and right now those projects are screwed as simple as that it's like unless they find a source of funding they will eventually die out and the re and
because of this the sort of the parent company that we we built for simple ble we call it the california open source company the overall mission here is not just to serve simple bly but to help other projects that are in this stage to go from like fully open source and not no resources whatsoever to go commercial to the point where at least some additional work can go into it and the developers can manifest their vision to its fullest potential. Interesting. So with open source, it seems like...
If it's a really niche product, then people want to, if they're giving their time for free, they want to know that this is important and that they're impacting lots of things. And if that's not the case, then...
Why bother? Whereas if you find a niche market that's willing to pay, then it's worth paying attention. And maybe the converse is true. If you've got something which... is free and lots of people are using it, then there's more motivation to keep on making it free because you're doing something of consequence with your life.
Yes, but it goes even beyond that. Because if you make a project that becomes really public, like millions and millions of users, you fall into a category that I call too big to fail. Basically, so many people depend on what you're doing, including big companies. that it's better for them to sponsor you because otherwise their own revenue stream will be affected. But in a case, I mean, Simple will be at least a good example. Sure, I mean, there are other ways to do this.
And they're not as painful, but there is value there, but it's not as great. And it also doesn't affect any big company's bottom line because otherwise, yeah, sure, they'll spend like a couple hundred more engineering hours on... on this feature instead of just purchasing our license but they can still do it and you end up in that kind of situation but when it comes to a small developer in a project it's not just the
oh, I want to see it matters. It's also the relationship you have with the people using it. Because, for example, before going commercial, I've had companies reach out to us asking, hey, I want this feature. Can you do it?
and when you tell them sure but it's i don't like 200 hours of work are you willing to pay for it you hear crickets it just goes to you they don't want to get involved or it's even worse sometimes they demand fixes or whatever like i this broke i mean this is breaking production why can't you fix it i'm like i don't but you don't work on this so there's also always like it's the the positive side from the user but it's also a very toxic component here and when you're
most of these projects have just one or maybe two developers and they're doing it on their free time and they spent maybe thousands of hours on this. It's very easy for a project like this to fall out of grace. even for life circumstances because oh i don't know uh they need to i don't know one developer wants to move to a bigger place so they need a big bigger income and they cannot use this time for that they have to like get a different job or something else
So it's, I think we're losing a lot of people to this situation. And there are lots of very small niche problems that you, that every time you sort of face them, you realize. this should be easier this can't be that hard when you're when you're developing on something and that's the sort of indication that okay something's going on there and sometimes solutions exist and like oh the approach was great
But they stopped maintaining it, I don't know, three years ago. And I cannot use it because technology has moved past. Yeah. That's sort of the saddest thing. And we really want to fix that. We really want to find a sort of balance. And again, it's not about... getting rich it's about like making sure these things are sustainable because in no other industry is free stuff pretty much right as with the software industry and sure for For certain companies, again,
like big companies sponsoring projects for example i think meta did a great job by open sourcing react and about and a lot of their internal tools are also open because now they've standardized whole industries around their own technologies which is great And that makes sense. But that same scheme doesn't apply for smaller projects because, I mean, sure, a couple people can learn simple BLE, but I'm not going to become an industry standard. It's not realistic.
Kevin, I've been really interested, as I've been learning a bit more about... you, your business. But one of the things that fascinated me was your connection with Argentina. So that's where you did you spend your entire early part of your life in that country? Or what's the story there? Yes, that's correct. So I was born there in 1991, so 30 plus years ago. I don't want to know the exact number at this point.
and yeah i did most i well i graduated high school and did my university studies there and then i moved to the us just before the pandemic so that was the i got to the us i think a month before lockdowns start on all that and it's been a very interesting journey to see everything that has happened since i would sort of call them humble beginnings yeah when when i was just a student and like dreaming of everything that
that's all the things that cool things that were happening all all over the world and to be able to get somehow involved into that it was uh i think it's a really big uh sort of thing that my my younger self would be very impressed Yeah, I grew up in England, as you can probably tell from my accent. And I kind of looked at the planes flying towards America and looked at all the American TV and saw all the inventions, the goings on in Silicon Valley.
And, you know, back when I first came over, the internet didn't exist. But there were the newspapers in Silicon Valley. And you can look at the adverts for... This is who Apple was recruiting. And just that, just being, I felt like, oh, man, I've arrived. So I imagine you must have had a similar thing of seeing this ecosystem from afar. What was that like for you?
Since I was young, I was always connected to the tech world in one way or another. So I would follow news coming out of here closely, but it's a very different experience. to actually living in that space to seeing it firsthand i mean it's it's still a weird thought to say oh i'm driving past that building oh facebook is there that's where they make instagram oh there's oracle
I don't know. I mean, all the tech companies are pretty much all here. And you recognize all the brands. And back then, they would seem like something that's so far away, like so out of place from where you're from. And to just... be there and see okay there's the building i can walk in and say hello to people and it's that i think is a sort of experience you don't really
you don't really get until you go through it. If you're born here and this is a common thing, you don't understand how much of a privileged position it is to... be close by to where things happen and all the for example right now with all the like autonomous robots and self-driving cars
You see them on the street. They're there. They're not just a video clip you see on your screen. They're there. You walk past. You see a Waymo go by. You see self-driving Teslas. You see robots. It's... that part is still i mean i'm mind-blowing and i've been here for four years at this point and almost five and it's still like it blows my mind Yeah, I heard someone say that some of the greatest film...
cinematographers are foreigners who then decamped to America. And the reason they were so good was that they still had a sense of wonder about the landscape and everything. You're seeing it for almost... from this childish perspective. And I certainly feel the same way. The palm trees never get old for me. Although Argentina is pretty... What's the...
What's the geography of Argentina like compared to California? Oh, so the weather here in California is more or less similar to where I'm from, the city of Buenos Aires. Although... down there it's a lot more humid both in the summer and the winter but generally like the temperature range is more or less the same there's no snow humidity is different but aside from that and also it's a lot sunnier here in california compared to
down there but geography wise Argentina has pretty much the same features as you would find in the US I mean there's both like you've got glaciers, deserts rainforest and everything in between so it's a it's a beautiful place to to go and visit yeah you can definitely like just by i mean it's i think it's the 10th biggest country in terms of matters of surface so
It's a very big place. It's very empty as well. The entire country has the population of California, but one third of that is concentrated in Buenos Aires and the metropolitan area around it. Once you leave that area, it's very sparse. There's a lot of nature. There's not much civilization going on. So you can drive for hours and hours. There's a famous route that goes towards the Andes Mountains.
where you're driving for I think around 10 to 12 hours through a desert. And the road just goes straight through the middle of the desert. And most accidents there happen because people get bored, fall asleep, and just like drive off the road. You have those sorts of environments where there's absolutely nothing for miles and miles and miles and miles. So was it difficult to get into the States or not? For me, I was...
Very lucky to find a company that was very interested in hiring me over. So short summary is a project that I worked on for my graduation that ended up becoming my first startup. Well, that project wound down in 2015. And by that time, I shelved everything. I had everything there. But I found a company working on the same thing in the 2020s. And I basically reached out to the CEO and CTO saying, hey.
I know exactly what you're building. I know how to build it. I have all this expertise. Bring me over. At that point, I was convinced I wanted to come to the US to live. basically they were like they saw the potential and they were like very eager to to bring me over and thanks to that they basically there are a few interesting visa programs especially when you're collaborating with universities that can
to hear fast. So what I was working on was I was working for them, but also helping as a, I forgot the exact name of the title, basically like guiding practical assignments for students for the UMass Boston. So both things were happening in parallel. UMass Boston is a great program to help startups in Massachusetts to recruit talent from abroad by basically co-sponsoring their visas and all that. It's a very...
It's a very interesting program, by the way. And that was basically my step in. And once I got in, I sort of moved along. What was the technical hook? What was it that your startup was working on? So my first startup was a company built around measuring electrical signals from the forum, from this area here.
to recognize hand gestures so the very basics we had was like you can move your hand up down open your hand or close your fist those would generate very specific electrical signals on the forum that we could measure and we would use that as a sort of input for computer phone video games etc that was the sort of vision that we had in
Interestingly, despite the technology being available for so long, there haven't been many commercial deployments of this. Even the ones that Meta, for example, has been pushing after purchasing all the IP from a product called Mayo. I forgot the company name, but all that is still somewhere around it. We also did a whole bunch of development around just recognizing gestures based on IMU. So just movement without any electricals.
signals and that is like sort of slowly coming in but I'm at this point I don't really know why this never happened there is a I know there's a company that I follow on LinkedIn that every once in a while they post like similar demos to the ones we were doing like almost a decade ago at this point. And it's still not a product. And honestly, I don't know why.
And to answer your question, the company was doing something very similar, but measuring electrical signals in the wrist instead of the forearm. But conceptually, everything was the same.
So this seems like this put you on a trajectory to work at Neuralink, right? Correct. I mean, I do have... Most of my professional experience has been around... embedded devices talking over bluetooth with some connection to biological signal processing because of this and that and that i think was sort of the the entry
the entry into Neuralink. Also for my second startup, I had a bunch of experience like dealing with more complicated problems, like lots of devices. As a matter of fact, the project I presented for when applying a Neuralink was...
how we solve the communication between our sensor gateways and all the sensor networks that we would deploy for this company. I'll go into details maybe a bit later on. And yeah, it was mostly, I would say the thing that... they liked was that I was a good generalist that wasn't afraid of jumping into any kind of difficult problem and was
versatile enough to figure out, okay, if I need to learn something or just enhance my skills to be able to solve a problem, I would easily be able to do it. For anyone who is willing to apply to any...
And your VLANs companies in general is as long as you're good at what you're doing, you're definitely getting it. That's the key point, I would say. Okay. So the secret is be good. Yes. You'd be surprised at how... how easy it sounds but how hard it is I mean I've interviewed people who had I don't know like 10-15 years of experience on top of my
years of experience. And sure, they were very good at a very narrow topic. But the moment you shifted the focus outside of that, they would just fall apart. And dealing with a very complicated device, as is a brain implant. You need to know about those topics. You need to understand about how RF works, how control systems work. Yeah.
I don't know. I forgot what else. I mean, signal processing in general. I mean, you're dealing with biological signals. Okay, so any sort of processing around that, good handling of statistics because some things, some algorithms are statistical in nature. So you need to determine, okay.
how what's the best way to do this and to understand physics of it because you're dealing with well electromagnetism heat it's all all those things uh matter so you need to you don't have to be totally proficient but you have to be I would call it conversational in all those topics. You need to know, you need to understand what the core concepts are, how to communicate, how to at least operate with them.
Very interesting. And so how did you get the job? You became aware of the company and you went to their website and applied or what was it? Yes. Surprisingly, yes. I knew nobody there. So this was like right in the middle of the pandemic, the height of all that. I had, at my past job in Boston, I was sort of, I fell a little bit.
i was liking the company with less once i basically once i joined internal politics were a bit too much for my liking and the company was shifting its focus away from biosignals to more let's say defense kind of oriented projects it wasn't that wasn't really my my thing i my i mean the reason i joined them initially was to help people so then the first
designed to help people with ALS to communicate so because usually they keep some sort of electrical activity it might not be enough to actually move the hand or some muscles but it's enough that you can actually detect it so that that was my initial attraction to a company you know hey this is using technology i have and we're doing like good we're trying to help people and when that was sort of like when the focus of the company was shifting away from that i realized okay i
It's not the thing I wanted to do. And a friend of mine actually told me, hey, Neuralink is recruiting embedded engineers. Their description of what they're looking for is very similar to what you did. So why don't you apply? And I just applied online. I mean, described everything. I went through the process, had a call with my boss, sent me the sort of practical assignment to do it at home, spent a full weekend doing the best I could out of that. It was like...
It ended up being like, I don't know how many thousand lines of codes, like I think 12 pages of documentation was like insane. But I really, I mean, I really wanted to get in. I really wanted to get in. And so it was that. Then a phone call with my boss's boss. Then they flew me over, more exercises, meeting a larger portion of the team, seeing sort of the soft part of the interview. How would you fit in with the culture?
And yeah, once they said, yeah, we like you, we'd like to come here. We figured out all the integration paperwork to get transferred over. And I started working for them just after the three little pics demo.
in September 2020. And what was that demo? I've seen the thing with people, which is amazing. This was the first public demo showing that they could capture... as signals from from a live animal and they did a live demo about it so showing hey here's a pig look and we the implant is connected to a sensory part that's connected to the snout so every time they they poop the snout that you could see the like the activity going on
And I was, unfortunately, wasn't able to attend because I wasn't an employee yet. I was already here in California when all this happened. But I started a couple days after. So I... I even missed a very big party they had afterwards that I'm still a little bit salty about. I'm not going to lie about it. But yeah, it was a great experience.
And what was the test? What did you have to do in the weekend? What was the assignment they gave you? It was actually very simple. Here's a temperature sensor. Write a driver for it. But the thing here is... The constraint is, okay, this temperature sensor is working in a safety critical scenario where hardware can fail in any way. How do you deal with that?
That's when you actually have to start thinking the problem through first principles. It's not just what's wrong with the software, but what could be wrong with the hardware, what kind of communication could be...
Basically, what kind of messages could go wrong? How could you detect them? What happens if a line goes up or stays up, stays down? How do you see it? How do you recognize it? What information does the hardware provide? And you have to like... diagram the whole, all possible failure modes and then say, okay, how can my software detect each one of them?
and build the whole driver around that so for example if you know you've if you detect that the system has crashed from what seems like a hardware failure at some point you should stop trying because you might cause further damage things like that it's that the test itself is not very difficult i mean any like fresh grad can definitely do it but what
Sets candidates apart is usually how far they can think along these lines. Amazing. Amazing. My first technical job was writing device drivers, actually. And I wrote a book about it. I wasn't actually very good at writing the device drivers, but I thought the book was pretty good. So what did you end up doing at Neuralink? What was your job?
My job there was embedded systems engineer for the implant. So my role took working on the code running within the implant and all the SDKs and tooling around it. because of the nature of this device there's a lot of testing required and we we did automate a lot of it and for that you need like a big like a big layer of software supporting all of that on the outside
The great thing is a lot of this also translated to the manufacturing line. So we basically were very embedded into any sorts of activity around the implant. And it was also a great experience because... i had the chance to be involved in pretty much every aspect of the lifetime of an implant from manufacturing to testing to even destructive testing those are fun
yeah just and you basically get involved and you learn about okay what are the use cases what's the infrastructure the scale of it as well because something that i don't know many people realize from like what what everything that goes into like testing a medical device especially something as novel as this is how much you need to test because the thing i described previously okay you have to come
find out all the failure modes and understand why they can happen, how they can happen, how can you detect it, how can you mitigate them. Now imagine that not just in a driver layer, but on a full device layer, on something that's very complex, multiple parts.
it's the same thing it's like okay there's some principles you can apply to make things uh safer by design for example i don't know the hardware being limited to like the hardware on its own cannot affect any damage because it doesn't have any way of producing damage or from the software layer to say okay i'm gonna assume that any any communication will be compromised so any any command i receive
I will determine if it's safe enough before executing. Stuff like that. There's that sort of thinking that goes into it. And of course, testing, because you have to make sure, okay, let's cause this failure and see if you're actually following that and what can happen.
plus also the scale of it. Because when you're building something like this, you want to make sure that even rare accidents or rare failure modes don't happen. And the only way to do it is to test lots and lots of devices because... Let's say you have a failure mode that happens, I don't know, one in 10,000. You have to test 10,000 devices to find that one, recognize, oh, this is a new failure mode, document it.
and apply all the fixes and mitigations across the entire chain. So that's more or less the kind of work that happened there. It's about, okay, we need these features, and we need to test them, and every problem that comes up, we need to have a solution for it.
And were you doing any work with Bluetooth as part of the project? I mean, yes. So, as I mean, as I said multiple times, the idea is for this, for the implant to become the Fitbit for the brain. And the main protocol that is... pretty much everywhere and lets you do lets you do that is bluetooth it's the only one that is power efficient
and has a decent enough throughput rate where these kinds of activities can happen. I mean, Wi-Fi would be great if you could stream such a volume of data. But if an input like that had Wi-Fi, probably the battery wouldn't last more than, I don't know, 5-10 minutes. But the frequency that Bluetooth works at is famous for not going through water or flesh. So how do you overcome that? Making sure the antenna points out, not in.
It's basically that. I was just tangentially involved in that, but the team that designed the antenna for the implant is just something out of this world. One of the smartest people I've ever met. And they do magic. Antenna design is one of those things where I'm not very familiar. and seeing what they can do and how all these things affect the result. It's almost magical at this point. Or alchemy.
yeah yeah williott's antenna design team and they were like uh i saw them as gods really and interesting quite gender diverse so there's something about good antenna design that seems to be a really good fit diverse teams. So what, if you had to put your finger on how the team got so much done i mean this is a common theme in in the projects that elon does is just kind of seeming to do the impossible and and okay maybe it's not exactly when it was promised but it's also
10 times faster than anyone else thought was possible. So what's your perspective? Having worked in one of those teams, how is it that these teams get so much done and achieve things that no one else can do? I think, I mean, it's a great question. I think that it's a combination of two factors. So one, there is the talent and sort of the skills of the person.
Elon would usually say is you have to hire people where a skill to ego ratio is a lot greater than one. So you want people that are very smart, very capable, but also easy to work with so they can recognize when they're wrong and fix their mistakes.
Because there's nothing worse than someone being very confidently wrong. And that's one. And the other one, I think, is personal accountability. The way I would describe it is, it's not that a certain team... is responsible for a certain component part whatever it's a certain person so if anything happens with that you know who to go to you know who to ask and that person is the one that will respond and it
make sure that because you know i'm responsible for this you make really damn sure that it works because otherwise okay you're the one who everybody's going to be looking like hey what's up I would say if those two bases are well covered, the rest sort of falls in line on its own because you avoid diluting responsibility and you have people actually capable and executing on those responsibilities.
I think that's pretty much it. There's no secret sauce aside from just finding good people and making sure they know what they're working on, which is something that in the past I've seen is like, it's very...
It doesn't normally happen. I mean, culturally, we tend to say, oh, it's, I don't know, it's a firmware team. It's the hardware team. And I don't know, let's ask someone. Someone shows up. Oh, I did not design this component. I don't know. You have to find someone else. They have to ask their boss.
Or maybe that person left or like all those things add up in terms of inefficiencies. And once you have this in place, execution just happens. It's wonderful to see. It's like we would have... the the sort of the cycle around this and early would be great because you would see oh we have this really big problem that we don't know how to solve and
everybody's focused on that thing and in two weeks it's a done deal it's like okay sure yeah we we got it we know how to we know what we need to do it's done mitigations everything is implemented and we move forward and that sort of continuous focus and improvement and going and going, that basically gets you that far. And then sort of... project methodology, sort of stand-up meetings, sprints? I mean, there's a bit of structure, but in the end, again, you're personally responsible.
for a certain part of the stack. You just make sure that the thing you're working on is important for the company goals, which you'll talk with your manager. And that's it. And your manager will basically communicate up and down to make sure that...
these priorities are correctly set and they're also communicated up saying okay this is our status this is what we're doing this is how things are going and that's it in terms of like how much structure there's there's not much there's not more structure than that So, and again, you know what you're doing. And by keeping things, by keeping like very small teams and very focused.
You don't need to have that much overhead in terms of communication because all of this happens. You can keep it in your mind. You can keep the entire state of your project in your mind. And you know, every time you go in the morning, you know what to do. You know where you're left, you know what to do, and you know what the plan is. You have a vision for what you're working on. And that sort of continuity is great for these kinds of things because you don't...
Again, it prevents the dilution of responsibility. Yeah. But how do you structure the project? Is there a Gantt chart? Are you kind of doing iterative stuff or status reports? How does that work? No, none of that. It was just one-on-ones with... with our manager and it was things revolved around hey we need this feature or we need to solve this problem and it would be like okay i'll start working on it i'll i don't know depending on how complex it was like
It's done either in a couple of days, maybe a couple of weeks, and it's just status reports. Like, hey, one way to look at it is how to, when it comes to defining priorities, is there was usually, there's usually a... a global priority and everything stems down from that. My time at Neuralink, it was, okay, we need to get to humans. So this means building a device that is human-grade and getting all the necessary approvals.
And that was the main directive. That was the only thing that mattered. And every decision and every step you took was filtered by that. So you know, okay, is this getting us closer to this? Yes or no? If no, okay, you should be doing something else. And if you can solve for that, if you can understand, okay, if for the company to, the next existential risk or the next goal is this, then it's very easy for everybody to know, oh.
how i'm contributing to that uh to the goal and if you're not if suddenly hey i'm not doing any if you're like hey nothing that i'm doing is helping achieve that goal uh okay It's a good opportunity to talk to your manager and say, hey, I'm not helping the company goals. So why don't we focus on something else? Which is, I mean, it's pretty much the same thing that the way he's done, he's manages all his companies. Like, okay, there's an overarching goal.
and everybody should decide their own goals based on that very interesting so did you get to meet him uh yes i was there for several meetings all hands so he would come by like every once in a while uh he's very tall i mean i'm i'm not particularly short or tall but he's like like more than a i think head and a half taller than me like oh man it's a big dude but yeah no i mean i would say he's um
As a boss, I would say he's one of the most reasonable ones I've had. He's a very non-bullshit kind of person. So it's like, okay, give me the facts. Let's end.
explain to me what what you're doing what the logic behind that is help point out hey have you thought about this have you tried that and based on that you'll i mean you'll provide the feedback and that was pretty much it i mean he he would set ambitious goals but he also knew he knew exactly how far along to place them so that they would still be reasonable it's like okay it's it's because i've had managers in the past that have said oh we need to do
x y z and it's like you're basically asking me to do two years worth of work in a week that's not gonna happen there's no way but here would be something like he would be asking okay this is a Let's say it's a four month long project that he's asking us to do in two. That kind of thing. And it's like, okay, yeah, it's hard. We're probably not going to get there in two.
but maybe we can do it in three and you push and you because you're aiming for two sure deadlines slip whatever but because you have that strong forcing function it actually gets done in three and because everybody And because, again, for all the things I mentioned before, you have very clear responsibilities of who's responsible of what. And because of that, there's no responsibility dilution. So nobody can sort of point fingers to someone else and say, oh, I didn't do it because.
this other team didn't do it it's like no no you know exactly who's doing everything and it's sort of again it put it makes sure that all the like everybody's pushing in the same direction and you get to and you get to that yes Very good. Well, I could talk to you all day about this, but we've got other parts of the conversation to do. And one of the most important ones is, of course, your three songs that have some meaning to you. So what was your first song?
So the first song, I forgot the actual, like the artist name, but it's a song by Odessa. That one is, that song is the... The cover song that Neerling used for the first promotional video. Yeah. So I was basically what happened. I mean, they put out this video. I forgot exactly when I think.
I think it was already out when I applied. And when I... And I remember, like, just those days before I joined, it's... I was playing that thing, that video and that music over and over because... uh coming from argentina ended up working at such a place was it was it would that would that was pretty much my past self's dreams. Like, okay, I made it. I got here. It's happening. It's, oh my God, it's happening.
And the sort of the big shift in my career that happened after that, because again, once you start working in such an environment, I can definitely say my engineering skills have improved a lot just by being surrounded by people. who have done this for long enough and teach you how to think this way. It's not something that you learn in college. So you really need to be exposed to this kind of environment.
and you do get and it it's for me it was sort of intoxicating at that point and it's uh it's a very clear uh for me it's a very clear reminder of the basically that You can just do things and even the things that seem insanely far away, you can actually get there. It can happen. It's a lot of work, but you can get there. And for me, it's that reminder. It can happen.
Very inspiring. Very inspiring. I love that. What a good choice. What about song number two? Song number two is something just like this. That one was this song. It's a sort of a... a bit more of a personal story, that's the song that made me realize I wanted to propose to my wife. Okay. I had been with her in a relationship at that point for... i think it was like six or seven years and i was on a trip and i was in a trip with some friends and they told me hey
I was there was basically some things going on in the relationship. And when I told him, hey, this is happening, the response was like, you're an idiot. She wants you to propose. Just do it. And I was so it was weird because I was so.
dumbstruck but they're like wait really that's what she wants i i missed that but completely like fully fully to the point that after i got by i basically apologized like okay now and once the whole thing was really i apologize okay now i understand sorry but it like wait if it's when you're like missing the hints but completely and when basically when i came to realization i heard that song i was thinking like
small hotel and i was having breakfast when that song was playing on on the tv like in the heading area and it sort of clicked uh if you look at the lyrics like what are you looking something just like this and it's sort of like it made that click it's the it was this uh it's this realization that oh you can actually like it's okay i really want to marry this person i want to
i want to go through with it and i'm gonna i want to make this commitment and again just blasting that song over and over like sort of reinforcing that feeling and all that and yeah we basically got engaged I think it was still a year after that. No. Wait. No, no, no, no. I got engaged a few months. No. My wife's going to kill me. We got engaged a year after that. Married a year.
a year after so two years after so it took some some time and yeah we've been married at this point it's going to be five years now in july and we now have a seven month old daughter so like we've come a long way since then And your wife's from Argentina as well? Correct. Yes, I met her in college through a common friend. And we've been together since I think it was our second year. So yeah, it's been a long, long journey.
Looking at the pictures of us when we were that young, it's like, oh man, time has gone by. Very good. And what's song number three? Song number three, it's Basket Case by Green Day, although I had a few others that could fit into that.
that role like um original prankster by the offspring and basically i like that song because it sort of symbolizes a bit of my sort of call it crazy nature in terms of oh I always want to do things and like it's always something happening and I'm also a little bit of an asshole so like a reminder of that is also not a bad thing every once in a while so I think those are like good reflections of my personality and I think they're a great way of showing what you or like expressing that sort of
teenage anxiety and like oh i want to try all these things but everything's hard everything's complicated but i'm still gonna do it i'm gonna mess up some people in the meantime like you just go and do it i think that sort of spirit is the thing i was trying to convey Well, you don't seem like a bad person. Why do you say that? The way my friends put it is I'm as subtle as a brick to the face.
I usually, I, this was something that sort of developed over the years, but I'm very, I learned to be very direct and I don't like when things are sugarcoated in any way. So, and instead of, if I think. You're doing something wrong. I'm just going to go and say it. I'm not going to sugarcoat it. And for example, I've been helping some startups as an incubator here at UC Berkeley that I collaborate with as a mentor.
And sometimes you have to tell participants or like potential entrepreneurs, hey, your idea sucks. You should think of something else. And I've had a few cases where I was the first person to actually tell them that.
because everyone's like no no your idea is great you should keep trying but you know they're like lying to them because you can tell from a mile off from a mile away the idea is bad so i i sort of doubled down on like brutal honesty and hey if you if i think you're doing something wrong i'm gonna tell it and some people like it some people don't and yeah fine when they get upset but
Yeah, for those, I'm definitely the asshole kind of person. Well, I have a startup I want to tell you about later, but we don't have time to talk about it now. Kevin, this has been wonderful. I've really enjoyed the conversation. Thanks very much for coming on the show. No, thank you very much. So that was my conversation with Kevin. Really fused core Mr. Beacon geeky tech discussions.
And the other side of Mr. Beacon, which is kind of bigger picture, what's going on in the world of startups. And I love it when both these things come together in the same conversation. So you're exceptional. You listened to the end. Thank you for doing that. Thank you to Aaron Hammock. for editing and pushing this out. Sierra Walden for doing her part in publicizing the show and getting it online.
Do let me know if you have any feedback, any requests. In the meantime, enjoy the journey. Stay safe and be nice to each other. All the best. You just realized your business needed to hire someone yesterday. How can you find amazing candidates fast? Easy. Just use Indeed. Stop struggling to get your job post seen on other job sites. With Indeed Sponsored Jobs, your post jumps to the top of the page for your relevant candidates so you can reach the people you want faster.
Sponsored jobs posted directly on Indeed have 45% more applications than non-sponsored jobs. Don't wait any longer. Speed up your hiring right now with Indeed. And listeners of this show will get a £100 sponsored job credit. To get your jobs more visibility at indeed.com slash P-O-D-K-A-T-Z 40. Just go to Indeed.com slash P-O-D-K-A-T-Z 14 right now and support our show by saying you heard about Indeed on this podcast. Terms and conditions apply. Hiring. Indeed is all you need.