ºÚÁÏÃÅ

Podcast Transcript

Zhamak Dehghani:

Hello. My name is Zhamak Dehghani, and I'm recording a session of our podcasts on hardware development with my cohost.

Alexey Boas:

Hello, my name is Alexey Boas. I'm the head of technology for Brazil, and I'm speaking from San Paolo right now.

Zhamak Dehghani:

Great. And I am in San Francisco. We have two of our passionate developers who've been working on hardware for quite a few years with us today from Brazil. We have!

Desiree Santos:

Hi, I'm Desiree, consulting developer.

Zhamak Dehghani:

And from Sydney, we have!

Charlie Gerard:

Hello, I'm Charlie, I'm a software developer, and I'm really excited to be talking about IOT today, so thanks for having me.

Zhamak Dehghani:

Thanks for coming. Yes. Desiree Santos has been a software developer for quite a while, but she has been quite passionately involved in hardware development since 2011, and she's been playing around with different tech stacks and different types of projects.

Zhamak Dehghani:

Anything from C++ and Rust and Raspberry Pi and different kind of hardware boards.

Zhamak Dehghani:

Charlie Gerard has an interesting perspective on IoT development because she's been kind of playing in the intersection of IoT design and art.

Zhamak Dehghani:

So we're super excited to talk to you guys about your experience today.

Alexey Boas:

And we are here at this time to talk a little bit about hardware development and internet of things. And so, there are a couple of different definitions, and there's also some confusion.

Alexey Boas:

So, internet of things. Hardware development, smart ecosystems.

Alexey Boas:

Desiree Santos, what's your view on what IOT is, and then what's involved when you talk about all these things?

Desiree Santos:

First of all, I love your question, right. But, I think that we have a few different terms of IOT definition, right?

Desiree Santos:

In my vision, and also understanding is we'll align it with... I can say we are garnered for being conceptualized definition.

Desiree Santos:

Imagine that the drink that you like, coffee or beer, it's over, right? And the refrigerators send a notification to your supermarket.

Desiree Santos:

That sends a box in delivery, your drink to your house, and debiting your account automatically. Right?

Desiree Santos:

It sounds like a dream. In the [inaudible 00:02:21] future. But it's not. It's real, and we're talking about the internet of things, right?

Desiree Santos:

And it's very interesting because when you're talking about the definition of internet of things, for me it means like a network of physical objects.

Desiree Santos:

So when you're talking about your drink, it's the physical. It's going to be like also your pet, your clothes, your refrigerator.

Desiree Santos:

These physical objects contain embedded technology, right? That could communicate and sense. It means get all information, getting the information from the environment, temperature, image.

Desiree Santos:

A lot of information and sends to somewhere. It could be local or in the cloud. And also interacting with the environment. The internal state or external environment.

Desiree Santos:

I think that it's a very important because when we are talking about internet of things, we have a different relation. It's not only people communicate with people. Like we have the internet today, right?

Desiree Santos:

But, we are talking about the people interact with things and things talk with things, also.

Desiree Santos:

So for me, the conception of internet of things, it's a mix of network, physical objects, communication, and sensor and exchange this data. To improve our life, increase the precision. For us to make decision.

Desiree Santos:

We want a good and better life. And I think that IoT really helps in part to improve this.

Alexey Boas:

That's great to hear, and thanks, Desiree Santos. That's a very good explanation. It definitely helps understand what this all means.

Alexey Boas:

And Charlie Gerard, what's your view? You've worked closely in the intersection of hardware and software and design. So what's your take on that? What does that mean in general?

Charlie Gerard:

Well, to me, IoT is to put it in a really small definition, I like to just say that it's like a giant network of connected things.

Charlie Gerard:

So, anything that can be connected to the internet via Bluetooth or WiFi can communicate together, and you can create new experiences either in, as Desiree would say, the kind of like automation of your life. Where you could have, I don't know, when you stop your alarm in the morning, your alarm is connected to your coffee machine and it starts your coffee straight away.

Charlie Gerard:

So, you have like a lot of different fields where IoT is present. But yes, where I like to actually... What I'm interested in is more in trying to create new interactions and new ways for people to interact with devices.

Charlie Gerard:

You have examples of, for example, dancers wearing certain sensors, and it interacts with the visualization in real-time. So depending on their movement, you have different colors or things like that.

Charlie Gerard:

So IoT can be used in really like a lot of different fields. I kind of like to try and explore, try to push the boundaries and really see where it can go.

Alexey Boas:

Oh, that's great. And what's some cool things you've been seeing lately, Charlie, you're working on and what's been exciting you related to that?

Charlie Gerard:

So what I've been working on for since I started being interested in that space actually is around motion sensors.

Charlie Gerard:

Either sensors that you wear on your body and you try to understand the natural movements of the body and apply that to certain actions.

Charlie Gerard:

For example, I was playing with an armband sensor called the Myo armband that senses the activity of the muscle in your arm, and you can detect if you're doing a swiping right movement or left.

Charlie Gerard:

So you can swipe your slides when you're talking about something, and you're doing a presentation, or I've been more recently playing with a brain sensor.

Charlie Gerard:

You wear the headset, and you can train it using machine learning to recognize certain mental commands, certain thoughts like pushing something.

Charlie Gerard:

And I was working around trying to build an interface where you could navigate using certain thoughts, which were still very early stages, but you also have facial recognition. When I'm looking right or looking left, I see different things in an interface. That's what I like to play with, really.

Alexey Boas:

And how about you, Desiree. What have you been playing with or seeing that's cool and excites you?

Desiree Santos:

Yeah, so I have a lot of projects to share and talk, but I think that I will start to talk with a project. Drone talk, IoT project from ºÚÁÏÃÅ, that is a silent [inaudible 00:07:00].

Desiree Santos:

It's amazing project because, since I'm from Brazil, I think the IoT market in Brazil is more regarding like the farming.

Desiree Santos:

And because of this, this project is calling so much of my attention, and I like to share. Right!

Desiree Santos:

The idea of this project is something like around the technology solution. Take the guesswork out of the milk production for farmers in UK. Right?

Desiree Santos:

And how it works, a cow wear a collar that contains a 3D solar meter, right, which transmits this data to a base station inside the farm.

Desiree Santos:

It's very interesting. It was called to my attention because sometimes you're talking about internet of things, always say the information to internet, but internet is, I think, instead of thinking internet, start to thinking the network. And sometimes like MESH. Because of a lot of constraints.

Desiree Santos:

Also, the security. Okay. But coming back to the project, it's interesting.

Desiree Santos:

The idea of this typical event, the cow, right, is the head. She moves a lot. Right? And also we have their fault.

Desiree Santos:

You can understand how much is moving and this movement. And all this movement are monitored to let the farmer know the right time, the best time to start the process.

Desiree Santos:

And also to improve the quantity of milk in the end, of course.

Desiree Santos:

The collar has the benefit of giving the farmers inside the hardware and maximize the amount of milk being produced.

Desiree Santos:

And I think it's very interesting because we have from the bottle, right? The 3D solar meter inside of the collar and sending this information to a base station.

Desiree Santos:

Of course, the main part is the data. Of course, the hardware is very interesting, but the data is the best part. Is the main part. Because of the data, we can increase our precision, and of course, when you're talking about the farmer, increase the amount of milk.

Desiree Santos:

It's a very interesting technology. And also it's a project from ºÚÁÏÃÅ, right?

Desiree Santos:

But, also, I have a personal project. Sometimes I like to create, build some projects. But sometimes I like to understand more, how can I bring some conception, from software development, and improve the internet of things process to the development process.

Desiree Santos:

And when you're talking about the continuous delivery, the continuous deployment in my day, I would like to do the same in internet of things project.

Desiree Santos:

And because of this, I'm starting to study some, "How can I improve this, how can I make sure that I could apply in a safe way?"

Desiree Santos:

And I decided to start some system operational that we call snappy, and this snappy is very interesting because you can start on the ground.

Desiree Santos:

I can talk about also using some [inaudible 00:10:39] board or another microprocessor, right?

Desiree Santos:

This operational system is very interesting because you can use this app, and this app is responsible to wrap and isolate your project. In the inside of this is embedded, right? An embedded system.

Desiree Santos:

And the other part is interesting. It's because this system. This Board, I actually brought, is connected with Apple store, not Apple store, but Apple application store on the cloud. Right?

Desiree Santos:

And every time that you have... You apply a new release. You automatically update all the projects.

Desiree Santos:

For me, it's interesting because I can move and apply and test how we can deliver in a fast and secure way when you're talking about 1000 hardware.

Zhamak Dehghani:

I like to kind of double click on some of these stories you shared and maybe talk about the technology behind them.

Zhamak Dehghani:

Charlie Gerard, you mentioned brain sensors and controlling the world with our brain.

Zhamak Dehghani:

That's scary and fun at the same time. Can you tell us a little bit more as how the brain sensor works? Especially, that calibration process that you mentioned and the tech stack behind it?

Desiree Santos:

Yes. So the brain... Since I've played around with two brain sensors. The first one was called Neurosky, but it doesn't have a lot of channels. So it doesn't really track. I wasn't really happy with the way it was tracking things.

Desiree Santos:

So I moved on to using another device called the Emotiv, and the Emotiv has 14 different sensors all around the head to be able to track different things.

Desiree Santos:

So the facial expressions and also some mental comments. And, when I started working with it a few years ago, the SDK was only in C++ and Java. So I wasn't really familiar with those technologies, cause when I learned to program, I learned Ruby and JavaScript.

Desiree Santos:

But I wanted to work on something that would allow more developers to play with that. So I built a framework in [inaudible 00:12:49] to be able to let JavaScript developers use the data and build JavaScript applications as well.

Desiree Santos:

So, yes, now in terms of tech stack, I think the Emotiv team has been working also on like a Python SDK, but at the time, it wasn't there.

Desiree Santos:

So, now there is C++, Java, Python and JavaScript. So the thing that I've been building, it's still like a work in progress.

Desiree Santos:

But it's really interesting to be able to get the data in nodes now, and either send it to the browser via whip circuits or interact with other devices or things like Arduino and Raspberry PI, all in JavaScript as well. Yeah, this is how it works.

Zhamak Dehghani:

Do you have a particular application in mind or a particular use case in mind for this brain sensor?

Charlie Gerard:

I built two little prototypes when I'm working with the device, and one of them is to have an interface with a keyboard. So you just have the letters of a keyboard but on a webpage.

Charlie Gerard:

And I'm using the movement of my eyes to be able to select letters and be able to write a message on a webpage just by using the movement of my eyes.

Charlie Gerard:

I actually started working with that because, well, that's a little bit of a personal story.

Charlie Gerard:

But, when my grandfather passed away, he actually was in a state where he couldn't move at all. And when I started playing with that brain sensor, I really realized that it could help people communicate with just the movement of the eyes.

Charlie Gerard:

Maybe not me, but it could help other people. And so that's also why I like to make it open sources.

Charlie Gerard:

If anybody is in the situation, they could actually buy the sensor and potentially use the framework that I built and let a member of the family of someone they know, be able to communicate with other people just by using the movement of their eyes, which is something that I think is really exciting.

Charlie Gerard:

But, otherwise, if we keep it to hardware, being able to just move things just by thinking about pushing something. So, you think about the action of pushing, and I can make a little robotic ball move forward, or a drone take off.

Charlie Gerard:

And it's not really useful. It's very much like experimentation, but it's quite extraordinary that now you can buy consumer products online that allows you to play with your brainwaves.

Charlie Gerard:

I mean, that sounds crazy to me, but it's just getting more and more accessible to a lot of people.

Zhamak Dehghani:

Yeah, I think that's such an important point because we often forget that hardware's been around for a long time. Connectivity to internet from hardware has been around for a really long time.

Zhamak Dehghani:

Suddenly, IOT becomes the thing, but it's that accessibility and the maker movement.

Zhamak Dehghani:

And, accessible hardware, schematics, prototypes, the whole maker movement pivoted, I think the industry and the ecosystem. So, that is super inspiring.

Desiree Santos:

But, one thing that you said is very interesting also, because I think that also become more popular because of the low cost of hardware.

Desiree Santos:

And also, when you're talking about the prototyping area, you can say the Arduino change drive it completely because it's open, easy to use, and very accessible.

Desiree Santos:

Even in Brazil or even different part of the world.

Charlie Gerard:

Yeah, I agree. I think that I probably wouldn't have been able to start experimenting with hardware if I didn't learn about Arduino.

Charlie Gerard:

I think it was... They made it look kind of so easy to get started with. Because I think at first it wasn't even developed for developers. It was actually created to help designers get into hardware and build, or artists and build art installations.

Charlie Gerard:

And I think that really made it look so accessible to me. Then I started looking into this space, and then I go into more complex boards or sensors. But I definitely did start with Arduino as well.

Desiree Santos:

Yeah. Every computer. And when [inaudible 00:16:56] created Arduino, the main idea was, "So I really like to empower people that really don't know programming and have any idea about electronic."

Desiree Santos:

I think that it's the magic. It's the magic.

Zhamak Dehghani:

Yes. It's a level of abstractions that just didn't exist in the past in an accessible way. They'll do better. I think in terms of... We've played a lot in the prototyping phase, but for commercial production-ready systems, we can still share more layer of abstractions, open libraries, open hardware, open schematics.

Zhamak Dehghani:

Desiree Santos, on the internet of cows project that you just shared with us, that was really fascinating because if I understood correctly, the farmers were using this hardware accelerators in the collar of the cows and then tracking their movement, and from the movements, they could then predict when the cow has most milk and be able to kind of utilize that information. Is that a correct understanding of that project?

Desiree Santos:

Exactly, You got the idea. Exactly. I think this project is amazing because also you can improve, maximize the production of milk, and now also detect some disease.

Desiree Santos:

Only because of a specific known movement from the collar. You can define this.

Desiree Santos:

And all this information I can say is data. Right? Data has a strong, strong power, and we really need to take care of it and increase our knowledge around this.

Desiree Santos:

Embedded data and all the data science. I think it's here. Yeah. This is the time. This is the era.

Zhamak Dehghani:

The interesting fact I had heard about that project in terms of the diseases was the cows actually get depressed.

Zhamak Dehghani:

And when they get depressed, they go far away from their herds and wander off, and maybe die.

Zhamak Dehghani:

So we can kind of get them back to life.

Desiree Santos:

Yeah. It's amazing because imagine a farm is very, very big and a lot of cows. But you know exactly the cow X and B has a disease or doing some more movements that you can improve for that cow. Internet of Things, it's fantastic.

Alexey Boas:

And this again, Charlie Gerard, you both have experience developing commercial software products, and working software consultants. What would be similarities and differences that you would highlight when you develop commercial software versus working in an IoT or hardware project?

Alexey Boas:

What are some of the main things that are different?

Alexey Boas:

Desiree Santos, you mentioned some of the deployment challenges and doing a distribution of binary's and then doing continuous delivery in that context. Are there other aspects that you think are relevant?

Desiree Santos:

Yeah, I think, of course, there are a lot of difference between one and the other, but I think that looking this perspective, when it's time to think about internet of things projects in more complex.

Desiree Santos:

Because you also need to understand where these projects will be. It's inside the house. It's outside the house, is a farm. And it's also the kind of sensor you need to use. You need to understand the best connectivity to use.

Desiree Santos:

It will be online/offline. Also, we have a different variety of states when you're talking about connective the machines, right?

Desiree Santos:

If the machine is WiFi off, it's on. If it's busy, it's sleeping.

Desiree Santos:

You need a lot of things to take care.

Desiree Santos:

But, when you're talking about also the software development, I think the main part, and a lot of the mistakes that I saw a lot, is when you're talking about some protocols, right?

Desiree Santos:

When you're talking about the interrupted ability. Because we are dealing with a lot of, and different kinds of hardware and sometimes hardware means we have some relation with the fabric.

Desiree Santos:

That you need to use a specific protocol and your application should deal with different and multiple protocols.

Desiree Santos:

And of course, for IoT, we need to take care about the size, about the restriction of the band, about the restriction of the hardware, and it guides us.

Desiree Santos:

Instead of using HTTP we usually use, we need to take care and think more about messaging.

Desiree Santos:

We can say protocol like MQTT or COOP or if we talking about MESH, maybe Zeeb or Telehealth.

Desiree Santos:

Something like this, but it's a different perspective. We need to take care of much more than regular or normal software development.

Alexey Boas:

And Charlie Gerard, I know you use a lot of creation process when working with hardware. So could you tell us a little bit about this prototyping cycle in that context, and then how it works and some of the highlights and pitfalls of doing this kind of stuff?

Charlie Gerard:

Yeah, so every time I come up with a new idea... So, for example, in Sydney, we moved offices a few months ago, and we wanted to start and build kind of our installation in the office that would involve using voice recognition to display words on a matrix of LEDs.

Charlie Gerard:

So that sounded really exciting, and I had an idea in mind, in theory, about how everything would work.

Charlie Gerard:

But then, when I started prototyping, I realized that it was not at all what I thought.

Charlie Gerard:

And that's what I really like about prototyping. It's you kind of think you have it all covered, but then as soon as you start, you realize that you're going to probably have to change everything.

Charlie Gerard:

And you iterate, and you start with boards like Arduino or Raspberry PI and you kind of attach sensors one by one, and you see what doesn't work.

Charlie Gerard:

And also, in terms of the amount of power that things are going to need.

Charlie Gerard:

For example, if you just prototype with an LED at first, you probably just need three or five volt or not even.

Charlie Gerard:

But for our matrix of LEDs, we started... We had to actually calculate how much one LED would need multiplied by, probably 1000, and then you actually deal with real power.

Charlie Gerard:

I haven't had the time to really keep working on that. But it was really interesting to start with Arduino for a very basic prototype. And then I realized, "Well, the Arduino can't drive that many LED strips."

Charlie Gerard:

So we moved to a raspberry PI, and then I was following a tutorial, and they were talking about having a FadeCandy driver between the Raspberry PI and the LED strips.

Charlie Gerard:

And then same again. I actually realized later on that the strips were too long to be driven by the FadeCandy driver.

Charlie Gerard:

So it was having all these different steps made me realize that prototyping is a very important stage. You learn so much about things, and then you start changing how the whole installation is going to work.

Charlie Gerard:

And yeah, so that's, that's my process really. Starting little by little, I haven't had the chance to put anything to production yet. So I don't exactly know what is involved later on.

Charlie Gerard:

But at least in terms of prototyping, that's definitely a stage that is important, and that you shouldn't forget.

Zhamak Dehghani:

I think that's what's really changed in kind of hardware development. Because even doing prototyping was such an expensive step. So you would normally traditionally have these very concrete schematics upfront and every iteration to produce the boards. And then every iteration of that will go through some sort of different hardware verifications, and it would be yet another multi-month expensive projects.

Zhamak Dehghani:

But, having an open hardware that you can just plug things to it and then at least get an idea if the idea you had in your head is doable at the prototype level.

Zhamak Dehghani:

Yes, production might have different restrictions and foreign factor restrictions, but even to be able to just touch and feel and get some feedback from the idea, that's super empowering. That we didn't have prior to open hardware movement.

Desiree Santos:

Yeah, I think it's the best benefit of a prototype, because during the prototype cycle, you can forget the best products.

Desiree Santos:

You can also focus the minimal of the product, like the MVP, the value to validate some things that are in your mind.

Desiree Santos:

When you're talking about this, the prototype for IOT, we can, of course, talking about Arduino. Because it's easy to use, and everyone can start from zero to hero.

Desiree Santos:

Doing your project. Yeah. And see things really happen. And no matter if it's the best products, the best sensor, the minimal you can do. And also it's very, very, cheap.

Alexey Boas:

Yeah. That takes us back to the things you were mentioning that made IOT and hardware development popular despite some of the things being available for so long.

Alexey Boas:

And, and Zhamak Dehghani, I think it was you that mentioned the maker movement as something quite important to make that happen. What's that? What's the big movement?

Zhamak Dehghani:

I think the [inaudible 00:26:48] for me is the accessibility and availability of the hardware to tinker.

Zhamak Dehghani:

Everybody who wanted to just tinker and build their own things. I think a few technical development made it possible. One is open hardware like Arduino and the libraries and their simple firmware that may not be production-ready, but it's good enough for prototyping, became available.

Zhamak Dehghani:

3D printer became kind of widely accessible, so people started making their own things, physical things much more easily. And there were businesses actually built and commercial offerings built to workshops to build workshops and allow people to go in and use different machinery to produce the physical objects and create the boards and so on.

Zhamak Dehghani:

I think what's interesting is how to replicate that within an enterprise and within an organization to create innovation. And through creating space for innovation at ºÚÁÏÃÅ, about five years ago, we started with a little global competition called 100 days of hardware, which was basically a global initiative for any developer who is interested in building something with hardware, to get his or her hands dirty to build the project.

Zhamak Dehghani:

We got support, financial support from the local offices, and we started investing in creating a makerspace or an innovation lab within every office. And I think at the end we had some sort of a prize for the winners.

Zhamak Dehghani:

But it just... That little nudge and that little support that we got from ºÚÁÏÃÅ internationally and across all the regions have created this wonderful spaces.

Zhamak Dehghani:

Now, every office has a maker space, has a passionate community around it that build things and also build capability to explore more commercially developed applications.

Zhamak Dehghani:

And I'm curious actually because I know more both Desiree Santos and Charlie Gerard are active members of their community. Maybe you can share your experience about the maker space and innovation lab in your office and what it offers.

Desiree Santos:

Sure. I love it. So when you talk about the makerspace, for me it's a magic word, right?

Desiree Santos:

Because if I'm interested about IoT, I learn in sign from the space.

Desiree Santos:

Makerspace and hackerspace, for me, it's like collaborative work, and we can build things together, and no matter if the person is from technology or not.

Desiree Santos:

Because sometimes inside the right hackerspace in Porto Alegre, we start coding in Rust or Go.

Desiree Santos:

And in the end of the day, we start to talk about the life, how life is.

Desiree Santos:

This is the point. As I reach a space to share, create and build the things together. And no matter what you know or don't know. I think it's amazing space.

Desiree Santos:

But the point is, this kind of space has tools that make you empower you to build and create the things.

Desiree Santos:

Like, do yourself, you can do. You can build and sometimes you can use learning to choose a 3D printer or help someone to build a smart plan, connect things.

Desiree Santos:

And I think this is a different mindset when you have it inside the office.

Charlie Gerard:

Yeah, no, I agree. I think it's actually really important to have like a makerspace. Because having a dedicated space makes you more likely to actually experiment.

Charlie Gerard:

Because you have this space where you have access to a 3D printer or different sensors, LED strips, anything that you want.

Charlie Gerard:

And this space is kind of only to build whatever you want and to experiment. And if the 3D printer makes a bit of noise, you're not bothering anybody else because you have this space where you can build.

Charlie Gerard:

And, also if you start building something, you can just leave it outside in this room, and people get curious, and they ask you a little bit about what you do, and maybe they even contribute with their own ideas.

Charlie Gerard:

And it's kind of creates a different dynamic where you can just let your creativity go, and you build whatever you want, and you actually have the space to do it.

Charlie Gerard:

You don't feel like you have to share a desk on a normal table where a lot of hardware can actually bother quiet people. Where it can get a bit messy. You can actually have this room where you can just borrow things whenever you want and experiment with it.

Desiree Santos:

Yeah. And because of this, I think the maker culture is very strong inside of this collaborative space, right?

Desiree Santos:

Because sometimes, every time they want you to start a project, the first thing is, "Okay, I need to learn A, B, C, D." And after it starts to work.

Desiree Santos:

In the makerspace, it's completely different. It's change this mindset.

Desiree Santos:

Instead of listing all things, it's changed, "Let's try to start with the minimum that we have."

Desiree Santos:

And it's a lesson for life., Not only to build things but for life.

Desiree Santos:

And it also changed a lot of my daily work because every time you're assigned a project is the same mindset.

Desiree Santos:

Or okay, "What's the minimum that we need."

Desiree Santos:

Not necessarily to have, the sandbox and the AWS or in different place. But, let's start doing a little bit in spite.

Desiree Santos:

I think this kind of mindset. It's changed a lot of our day.

Zhamak Dehghani:

That, I think, was quite evident when I went back to Sydney last week.

Zhamak Dehghani:

After a while, I noticed the difference between makerspace in Sydney when we first started.

Zhamak Dehghani:

When I first started setting it up five years ago, it was out of a few drawers in the middle of the office. So we had a few Arduino boxes and some sensors and other things in a drawer, and we will bring them out, put them in the kitchen and use the kitchen temporarily as a makerspace.

Zhamak Dehghani:

But, now, as Charlie said, we have a dedicated space with a lot of gadgets and goodies. Really creating an environment for learning, for experimenting. As you Desiree mentioned as well, just providing the conditions for learning, and this grass-root movement just blossomed itself.

Alexey Boas:

Yeah, it's cool. And on top of that, every ºÚÁÏÃÅ office is always hosting events and many things open for the community.

Alexey Boas:

If you listened to this and happened to be in Porto Alegre or Sydney, check it out. You might be able to get to know one of these spaces.

Alexey Boas:

So that's really awesome.

Alexey Boas:

Okay, so I guess we're coming to the end of this episode. It was a great conversation. It was great to have you with us. Thank you very much for joining. Bye.

Rebecca Parsons:

And next time, we will be joined by Camilla Crispim and Marco Valtas, both from Brazil, who will talk along with Neil and myself on how we make the Radar twice a year. So please join us.