The LED is flashing. The fluid is flowing. The latch is lifting.
This is how many people imagine the result of programming hardware. Something lightweight to the point of being uninteresting.
But what about an electron microscope? An ATM? Or the software in your car?
Does hardware recovery still seem like such a trifle to you? No, isn't it? And yet it's a far less visible part of the IT industry, at least to me.
That's why I invited Radoslav Slovák, the owner of Edhouse, who has been reviving physical things since the beginning of his business, in front of the camera and microphone. He has built a company that can now offer its clients an outsourced research and development centre. And it couldn't have happened just like that. That's why we asked the following questions.
🔸 What is specific about SW development for HW?
🔸 What is the competition in HW development?
🔸 How to build trust with a corporate client?
🔸 Will everyone have to know how to program?
🔸 Why should we want more women in IT?
TRANSCRIPT OF THE INTERVIEW
Martin Hurych
Hello. I'm Martin Hurych and this is Zahžeh. Today's Ignition is going to be about a little bit different IT and when I mean a little bit different, it's going to be about what might not be as visible and for a lot of you not as sexy, with Radoslav Slovak. Hello.
Radoslav Slovák
Hello.
What do salsa dancing and hardware recovery have in common?
Martin Hurych
Radoslav is the owner of Edhouse. Tell me what hardware recovery and salsa dancing have in common.
Radoslav Slovák
Absolutely nothing, but one great peculiarity is that in our group in Zlín, where we dance salsa, there are three IT company owners. So I don't think it's the hardware that's being revived, but there has to be some kind of correlation between IT and salsa. Don't ask me why.
How did Radoslav get to a company of 140 people?
Martin Hurych
It's weird. Maybe you're a bit of a hothead in Zlin. Let's take a brief look at how you went from programming in elementary school to a company of 140 people, so we know from what position you'll be commenting on what we're going to discuss here.
Radoslav Slovák
It's just that I knew I wanted to be a programmer since I was a kid. I studied it, as you mentioned, from elementary school, then college, math and physics faculty, and then some internship. Then, for family reasons, I found myself in Zlín, where it was very easy because there were no interesting jobs. There was no employer who would have been interested in me, who I would have wanted to work for at that time, so together with my brother-in-law Pavel Stržínek we founded Edhouse and gradually we set the goal to bring those interesting projects to Zlín.
Martin Hurych
How would things have turned out if your wife hadn't had a sister?
Radoslav Slovák
I can't say. I'm sure it helped, because we still operate in that ownership tandem to this day and it's a very workable model for the two of us, even though many people discouraged us and said tandem businesses don't work. I guess we are an exception to that, or some other case.
Martin Hurych
Congratulations, because they say the number of owners should be odd and three is too many, so congratulations on your success. When I read the preparation, one thing caught my eye. I will admit that in the fifth grade of elementary school, the most I wanted to be was either president or a garbage man. You started programming and you went for it. Do you remember what the first thing you programmed was?
Radoslav Slovák
I don't know if it was the first one, but the one I remember was back at the old Didaktik Alfa, where our elementary school teacher taught us programming. I think it was some simple pictures, we drew in basic with lines, I don't know if it was a duck or just some picture.
Martin Hurych
Is there anything that can be transferred to today? Has anything held up over the years in IT?
Radoslav Slovák
I don't quite know. I think I definitely have the DNA of an IT guy, or a programmer, a developer. We don't really like to be called "IT people", you know. I think it's that genesis, that gradual evolution, and you can see it in the employees that work for me on our team, that they're often people that have been doing it for a long time and you can see it in the thinking and the problem solving.
Martin Hurych
How do I tell my dad not to spill hot milk on me?
Radoslav Slovák
They are different professions. We are inside our bubble either developers, programmers, testers, SQA, project managers, which are obviously different roles.
What led Radoslav to skin HW? What does he enjoy about it?
Martin Hurych
I understand. I was saying that this episode is going to be a little bit about different IT because the bubble that's around me a lot is mostly custom applications, information systems, web applications and so on. You're going down a different path, you're bringing tangible things to life.
Radoslav Slovák
It's true. I guess the reason is that I personally had expertise from a previous job. So it was something that I understood and worked with, and quite logically we found our first customers in it. Then, like in every other business, you repeat that success where you've had it once before, where you understand it, and the domain of that hardware device application software put itself out there.
Martin Hurych
There still has to be something deeper, because if you weren't enjoying it on that first app, you'd run away from it.
Radoslav Slovák
All programmers and developers who work with hardware sometimes look down on people who do web development, databases and so on, because you can touch the hardware. It's doing something, it's blinking somewhere, it's running, it's moving, it's lit up, and it's a completely different feeling than just somehow fiddling with zeros and ones in your computer's memory or on a disk somewhere.
Martin Hurych
So can we give a typical application for the audience and listeners to get an idea of what you do?
Radoslav Slovák
Our biggest client is Thermo Fisher Scientific, a global company with an important branch in Brno, and we develop application software for electron microscope control. Electron microscope, lots of hardware, lots of construction, lots of boards, and our domain is in it. management and development of application software for such equipment. This is one example.
Martin Hurych
I've come across the view that hardware programming isn't that sexy because it can sometimes be simple, older chips, car apps and so on. You said yourself that in quotes just something flashing somewhere. I guess we're not seeing your full reality.
Radoslav Slovák
It is the case that many people think of single-chip development, some embedded development inside some single-chip in a simple C. There's obviously a great need for that, but it's really a minority these days because the problems that you're solving today and the ecosystem of those applications is much richer. Quite frankly, you need at most 10 developers for that embedded, for example, for an electron microscope. The applications around that, which are already driving on the PC, the web interface outwards, and finally the databases, because you have to store those results somewhere, are the rest. So embedded is kind of the tip of the iceberg that sticks out, but that whole ecosystem of absolutely all applications is maybe 95% or 90% of the problems that you're solving.
What does the competition look like in HW?
Martin Hurych
When an amateur looks around the IT field, he sees a bunch of information companies or technology companies in Prague, Brno and all of them are drawn to fintech, automotive and information systems. There is a lot of competition there. What does the competition look like in terms of hardware recovery in your field?
Radoslav Slovák
It is true that there is competition here too, of course, there is probably no business sector where there is not, but we are quite successful in finding interesting clients. Plus we are able to retain them and convince them of that through good work and what we do. We've been doing this for 16 years, so we know how to ultimately get the job done, get it developed, get it started, what the process is, the communication. The competition is there, but I always tell our new hires that we're looking for companies that have a package of work that they're able to monetize in some size 100. They have 20 of their own people for that package of work and the rest is a terrible field where you can develop something for those companies and they don't really know how. I think there's relatively a lot of that space. Of course it requires a very high level of expertise and knowledge of the subject matter.
What does the delivery to the client look like?
Martin Hurych
When we were talking before the podcast right here about this, I mentioned that you do turnkey development. It's gotten very hairy with you, so let me tell you what the delivery actually looks like or how we're supposed to understand it.
Radoslav Slovák
It's so diverse. Basically, we're only now realizing in hindsight that it really depends on how long you've been working with the client. Some initial input to that client could be likened to a simple remote body shop. You deliver your people with some knowledge of maybe a programming language, or some testing methodologies, project management. They have to come in to that client and the management, the work assignment has to be more or less up to that client. But what our value add is and where we're moving and what our goal is and what we're doing is to gradually learn that domain knowledge of that client. To learn to navigate that environment, to start forming smaller organized teams that take on more and more competencies, and that ultimate goal is to eventually take complete responsibility for a software release or hardware. So it's kind of a gradual process. If you ask me how we do it, we actually do it at the lowest scale of complexity of that management. It's more or less units of people that could be likened to a body shop, and at the highest scale it's really that complex turnkey delivery that you're talking about. Now you're basically looking with that client to see where the right threshold is to put that service of yours. There are other dependencies, another dependency is the availability of the hardware, if the hardware is not even in the Czech Republic, you can't take responsibility for some QA or some final integration testing. So it's very diverse.
Martin Hurych
Does that mean that the most that can be achieved is an external RnD center?
Radoslav Slovák
JIt's true. In this model, I'm proud that a few of those jobs, or some of those projects that we handle for some of those customers, are already working for me.
What to watch out for in the beginning?
Martin Hurych
If you were to start a company again, or had advice for someone who wanted to get into this business because they enjoy bringing things to life, what should you look out for in the very beginning?
Radoslav Slovák
That's a very difficult question. I don't like to give that kind of rakish advice. The market has also moved on a lot, it's been 16 years since we started.
Martin Hurych
So if you had handled what you know today yourself 16 years ago, what would you be more careful about today?
Radoslav Slovák
I don't regret anything that was done in that company, I actually enjoyed it and I don't blame myself. There are fouls and you have to accept them in business, but maybe one piece of advice, persevere. This is not an industry where you can expect any quick wins. It's really a long run here, it's a marathon, it's not a sprint, it's not a hundred. It's a marathon and you have to wait for it. I don't know if it's a good idea to skip any stages. If, for example, right at the beginning I don't know anything about the client, I don't know the relationships, I don't know the domain know-how, it's not good to say I'm going to cover the whole thing for him. That's a bit of a road to hell, so to speak.
What is specific about SW development for HW?
Martin Hurych
There was one other thing that caught my eye. Maybe it's my distorted vision, but for me, historically the creation of anything that is tangible is slower than the creation of software. Especially iterations are much slower. I remember a time when I was selling to manufacturing companies and from prototype to prototype was still calculated in months if not years in the absence of 3D printers. Does that mean that even the creation of that software has to be project managed in a completely different way, that you take a while and then wait a long time before something happens again?
Radoslav Slovák
It's true. Anyone who works with hardware knows the specifics here. The kind of slightly annoying thing when you're doing the software part in that hardware development is that you're basically at the end. You spend a long time sketching something, drawing it, designing it, then putting it together, dealing with the electronic design, making the first prototypes, and the software is the last thing. If there's any slippage, it's usually the software that gets all the slaps. But it has a way out. When you do really big orders, you don't wait for a prototype. You actually start developing the software at the very beginning, when the product is being defined. What's specific here, you have to create some emulations, simulators, and you don't develop against real hardware for an awfully long time. You're developing against, for example, an emulation that you built at the beginning. Of course, this is an additional added cost, but there is no other way. If you get the prototype and now you have to start embroidering the software, the delay is horrendously long. So these are the specifics of the procedure. The turnaround time in hardware is quite long, so quite often if the hardware doesn't do what
has, it's fixed in software because it's significantly cheaper and faster, if it can of course. These are all the specifics and you have to factor that into the project management of those projects and contracts.
Which clients does Edhouse specialise in?
Martin Hurych
We mentioned one particular application, the electron microscope. So what is the domain you primarily want to get into?
Radoslav Slovák
For us, the client is ideally a large company that has a proprietary hardware product that requires a fairly large and complex software base to function. Most of our clients also have their own development teams, they just aren't able to cover the full range of applications they need in-house. That's where we come into play. We can establish that relationship and we can create an external development team for that client. Most of the time they are multinationals, most of our clients have headquarters in the US or somewhere in Western Europe, so then we can also handle those things around that, getting under some kind of corporate supplier relationship. These are long term things, you're dealing with a contract for maybe three years.
Martin Hurych
So what do you enjoy segmenting?
Radoslav Slovák
We enjoy science, technology, and now we're looking a bit at projects in space. Personally, I'd love to have a piece of code flying over our heads somewhere in space beyond Earth's gravity. So we are looking at whether we are able to enter into some European Space Agency projects, for example. There, of course, the barrier is the administration, but it's there. Then it could be the medical segment, where there are also interesting pacemakers, pumps for diabetics, devices that need software and are quite challenging to develop.
How to build trust with a corporate client?
Martin Hurych
A lot of people around me get hives at the word corporate or automotive. It seems to me at first glance that this is an even harder segment than automotive. So how do you build trust or get business from this extremely difficult segment?
Radoslav Slovák
I listened to your previous podcast this morning where your speaker talked about this very thing and I just confirm his words. It's about people, it's about finding the right people in those corporate waters and being able to reach them. What a lot of people think is that once you get into that corporate, that's basically a gold card and you're out of the water. That's not true. You have to do a 110% job for them because they're going to replace you otherwise, that's the way it is. So it's that quality work, it's finding the right people and being able to create that space. We've learned that from our clients and we can replicate that success with others.
Is it better to understand everything and then delegate or hire an expert straight away?
Martin Hurych
Turn the page. When I visited you in Zlín, you mentioned several times in the preparation that you had a lot of hats on your head during the history of the company. You wanted to learn everything yourself first and then pass it on. The theory goes against that, take an expert, he will help you accelerate faster. I would be interested in your personal opinion as to why you have it laid out the way you have it laid out. Understand what's going on, surrender and be able to manage it. Where does this worldview come from?
Radoslav Slovák
I don't quite know, but my partner Pavel and I jumped straight into this business and we were both experts. Usually that tandem, when two people start a company, one is an engineer, one is a salesman, and then they usually end up having a fight. We were both techies with an analytical mindset, so we thought we'd try this route. In the beginning, you needed at least some administration, then you dealt with recruiting because we grew from two men in the beginning to today. The first 50 or so people we recruited ourselves and we had to learn and do it well. Only then did we find the professionals that you say are doing it better than we are today, and I'm glad for that. But that's a little bit off the point of your question. I don't know why we did it that way, maybe it was our inner disposition that we are the analytical types and we want to grasp it first, understand it and then pass it on. I don't think I can answer that any better.
What went wrong in Edhous? And what is the lesson?
Martin Hurych
You weren't always completely into hardware. People love to learn from other people's mistakes. Come tell us about it. I'm talking about the mobile app framework.
Radoslav Slovák
It was sometime around 2010, 2011. We were constantly flirting with the idea of making some product of our own. That's kind of the golden grail of anyone who starts a company, let's make a product, we'll develop it beautifully, it'll sell itself and we won't worry about the business. This thing came about a bit of a crisis because we had some customers quit and we didn't want to fire people. We had a product idea in our plate, so we had them embroider on it, program it, but the success didn't come there at all. We couldn't sell it at all and it was money wasted.
Martin Hurych
How big a step away from what you were doing at the time?
Radoslav Slovák
It wasn't quite that far, because in the beginning we weren't so hardware-oriented. We also tried different mobile apps, which was very trendy in 2010, and we didn't get to where we are today until later. So it wasn't completely out of the way, we knew the area, we were developing some apps in that area as well, but in this case we got burned on the fact that we underestimated that we couldn't sell a licensing-based product. Because it's about building a marketing team, it's about building a sales network, it's about building a brand, and those are skills that we don't know how to do. We can do development, that's our hoof, but we can't do end client sales.
Martin Hurych
Looking back critically, is it really that we couldn't sell it, or was there already a lack of market fit?
Radoslav Slovák
I guess you can't separate the two, they are always related of course.
Martin Hurych
Why I ask this question is because on the one hand, I see and am very supportive of people who want to break through globally with a product. On the other hand, they say, and it's been said here on the podcast as well, that there's a solution or an app for everything today. In hindsight, can you say what would you do differently today, what lessons have you learned so that the fuck up could be discovered earlier, or on the other hand, finish it?
Radoslav Slovák
It wasn't so bloody that it cost us the life of the company. I'm not going to say how much money we lost here, because my wife probably wouldn't have complimented me, but it's been a long time. Not to escape your question, I don't know, rather the lesson we learned from this is that we actually forbade it. We told ourselves we're good at this here, we understand this, we can do the production line to develop this software, we can approach this client, take the work from them and deliver it to them with a smiley face. So why would we do anything else and divert that activity somewhere else.
What or who is Edison? And what does it help?
Martin Hurych
I understand. Last topic today, you gave birth to Edison. What or who is Edison?
Radoslav Slovák
That's kind of our non-profit activity. Edison is a small educational robot, thanks to which we are trying to help primary schools in the Zlín and Olomouc regions in teaching children programming and algorithmization. We started this project in 2019 or 2018 and it is that we organize regular workshops for computer science teachers in our company once every six months. We're introducing them to one particular platform that they can use to teach kids to code, and I think it's working for us. We've had great feedback on it and we've done about 7 runs of these workshops and we've trained 160 teachers. I think it has a great effect because they are not afraid of the teaching, they use the robot, they integrate it into the teaching and I believe that in the Zlín and Olomouc region it will create a big wave of interest in programming. I hope that in 10 or 15 years we will have a second Silicon Valley there.
Martin Hurych
I really wish you would. Are we going back to where I started, for example, in Karl programming, or in basic programming?
Radoslav Slovák
Yeah, except that Karel is physical, the kids can touch it, they can make the program in a simple scratch programming environment, they load it into the robot and it does something for them. It's amazing, and the kids are having so much fun. I don't know how much fun you had with Charles.
Will everyone have to know how to program?
Martin Hurych
He went from right to left, up, down, and after a few hours we got tired of it, that's clear. I got two things in there. You mentioned teaching programming and teaching algorithmization. They're pretty far apart for me. Is the generation that you're trying to attract to programming today really going to have to know how to program every single one of them, like we have to know English now?
Radoslav Slovák
I don't think so. I see it as another addition to the general education of children. Consider that you also studied music education in elementary school and you're not a virtuoso, you took art, you painted a house and you're not Picasso. It will be the same in this area. I think it's good for the kids to get a little bit of a feel for it, to see what it's about, and maybe it's more about getting those potential kids excited about it and going after it. It's not like they're all going to be programmers now, I think that's misguided. It's just another addition to that education, the way we see all the other subjects.
Martin Hurych
I just separated programming and algorithmization on purpose because Karel gave me a completely different perspective on problem solving and algorithmization. There, for example, I see in my son that he is completely lacking and not learning enough. I won't learn to program unless I know the basics of an algorithm or can analytically write the process I am supposed to describe. Is that right?
Radoslav Slovák
That's right. I think this direction is important so that these kids learn to decompose problems into smaller parts, so that they learn to think a little bit in some procedural terms. When we started the activity here, we tried out the teaching ourselves on a group of children where we did some 10 lessons in a primary school. The feedback from the kids was that, for example, the boys told us that it was harder than they thought it would be, that they thought it would be a piece of cake, but it really wasn't. Even the simple problems we solved with the little robots in the artificial world were relatively complex. They said they couldn't imagine how your dad's car is programmed to park itself for you. So it's also the realization that it's not a given that these things work in the real world.
Martin Hurych
What did the girls say?
Radoslav Slovák
It was really interesting because the director told us after the lessons that she had no doubt that we would impress the boys. But there were also 14, 15 year old girls, I have one of those at home, who are usually hard to impress. But they were having fun with it, immersing themselves in it and creating. There was, for example, a very clear difference between the female and male element of thinking, but that's maybe a longer story.
Why should we want more women in IT?
Martin Hurych
Let's do it. I found out on your website that you have a special section for ladies.
Radoslav Slovák
That's right. Unfortunately, there are not many of them in IT and I think it's a mistake, because in the end, everybody uses those applications and it's 50/50, but it's mostly guys programming and designing. So there's kind of a clear imbalance there. Women are much more careful and they don't go into a problem without thinking, they want to be perfect and they don't want to make mistakes. But that's just the big barrier that we can't get them to do IT because guys are completely vain, but they go for it, they just try brute force. Now they're going to make 10 mistakes, but that's actually going to get them to the solution. If you suddenly want to be too perfect and you want to get it done right the first time and no mistakes, that's not fine. So that's the difference, in my opinion, where the computer science education should be encouraging those kids. We should be telling them not to be afraid to make mistakes, praising them for them and using them as an example. If you write a dictation, get two mistakes and get a B or a C, then those kids are afraid.
Martin Hurych
I understand that. Do I know a code written by a lady and a guy?
Radoslav Slovák
I hate to say it, but I believe the ladies really are more precise and meticulous. I don't like to introduce any gender stereotypes here, but unfortunately, I'm going to give an awful lot of them to IT and our industry.
What should those who would like to join Edison do?
Martin Hurych
Going back to Edison with my last question, what should those who are excited about this and might want to join you, help you, or copy you in another region do?
Radoslav Slovák
It's just that we do the training only for the Zlín and Olomouc regions, it's available on our website edhouse.cz/edison. We are currently developing a similar activity in Brno, I think the website is edisondoskol.cz. If someone else is interested in starting this activity here in Prague, they can write to me, contact me directly, they can come to our workshop to see how it works, I will be happy to explain and show them the whole workshop. We really see it as a volunteer activity and I think that if it is in more places, it will only be better for the Czech Republic, our IT and our future.
Martin Hurych
I'll keep my fingers crossed for you, because education needs to be shaken up. Thank you for your participation, may you thrive in both business and education.
Radoslav Slovák
Thank you, too.
Martin Hurych
We've had another episode of Ignition. I hope we've opened the windows again to a slightly different segment that we haven't had here on Zazh yet. I sincerely hope you enjoyed the episode, if you really did, make sure to like, share, talk about Zážeh with your friends, acquaintances, because otherwise the world won't know about us. Check out www.martinhurych.com/zazeh, where you can also find other episodes.I have nothing to do but keep my fingers crossed and wish you success, thanks.