How do you build a successful IoT solution? Peter Tuszynski and Przemysław Łagód of Intent join Ryan Chacon on the IoT For All Podcast to discuss building and scaling successful IoT products. They cover what to focus on first when building an IoT solution, how to approach finding market fit, hardware vs software, why some deployments fail, how to future-proof your product, and how to scale.

About Peter

Peter Tuszynski is a seasoned mobile software engineer based in Warsaw, Poland. He currently works as Chief Technology Officer at Intent, a software company building digital products at the intersection of physical and digital worlds. He’s been working with mobile apps since the days of Symbian and PalmOS. Nowadays he focuses on iOS and Android with a deep passion for machine learning and voice UIs. He’s the founder of a venture backed startup and former entrepreneur-in-residence at a leading VC fund, giving him a unique perspective of both sides of the table.

Interested in connecting with Peter? Reach out on LinkedIn!

About Przemysław

Przemysław Łagód is a highly skilled electronic designer and embedded developer with extensive experience in designing and producing electronic devices from scratch. With a passion for electronics and technology, he has dedicated his career to developing innovative solutions that meet the needs of clients across various industries.

Interested in connecting with Przemysław? Reach out on LinkedIn!

About Intent

Intent has over 10 years of experience conceptualizing, designing, and building apps for connected devices. They act as specialized think partners for products at the intersection of physical and digital - from net-new to products looking to enter a competitive and established niche. Their team has a proven track record of building and deploying groundbreaking digital products allowing them to quickly assemble and scale project teams from their talented pool of developers, engineers, QA specialists, UI/UX designers, and project managers. Some of their clients include Oura, BOSE, and McIntosh audio systems.

Key Questions and Topics from this Episode:

(00:58) Introduction to Peter, Przemysław, and Intent

(01:56) What to focus on first when building an IoT solution

(04:55) How to approach finding market fit

(06:16) Hardware vs software

(07:56) Why do some deployments fail?

(12:01) How to future-proof your product

(16:12) Is there a way to ensure product success?

(19:12) How to scale

(22:21) What are you excited about in the industry going forward?

(24:42) Learn more and follow up

Transcript:

- [Ryan] Hello everyone and welcome to another episode of the IoT For All Podcast. I'm Ryan Chacon, and on today's episode, we're going to talk about what you should focus on first when building an IoT solution, where most people fail, how to future-proof, hardware, software, those kinds of things, how to ensure or give yourself a better chance of success. And then how to scale from, you know, a few devices to hundreds to thousands and beyond. Very exciting conversation. I have the CTO and the Head of Hardware from Intent. They are a company that has over 10 years of experience helping to conceptualize, design, and build apps for connected devices. Great conversation. I think you'll get a lot of value out of it, but before we get into this episode, we'd truly appreciate it if you would give this video a thumbs up, subscribe to our channel if you have not done so already, and hit that bell icon so you get the latest episodes as soon as they are out. But other than that, on to the episode.

Welcome to the IoT For All Podcast. Thanks for being here this week.

- [Peter] Great to be here.

- [Ryan] Yeah, yeah. It's great to have both of you here. Um, very interesting conversation I'm ready, I'm excited to have here with you both. Um, but let me kick this off by having you give a quick introduction about yourself and the company to our audience and Peter, we'll start with you.

- [Peter] Sure. Uh, so my name is Peter. Uh, I'm a mobile engineer, uh, by passion and by education. Uh, I started building apps way before the term apps was coined. So in the days of Palm OS and BeOS, um, all that jazz. And I work at Intent as a CTO, and Intent is a, um, software consultancy that builds software at the intersection of physical and digital worlds.

So any project that has a physical manifestation is kind of our sweet spot.

- [Przemyslaw] My name is Przemysław Łagód. I'm Head of Hardware at Intent. My biggest experience is related to the hardware and embedded software.

- [Ryan] Fantastic. Alright, so I wanted to jump quick right into the conversation here. Um, a lot of interesting stuff to cover for sure. Um, the first thing I wanted to ask you from both of your all's vantage points, experiences, when it comes to building a solution, um, a lot of our audiences out there looking to adopt solutions, looking to bring solutions to market.

There's a lot of components that go into a successful solution from the hardware to the software, embedded systems, the cloud, the connectivity, you name it. How do you think through all of that? And what do you recommend that they focus on first, given they have kind of the choice?

- [Peter] So I this is probably the, not the answer everybody wants to hear, but I'll say it anyway. Uh, I think having a perspective, a helicopter view of everything, of every single component out there. It doesn't mean that you have to do it all by yourself in-house, but being able to holistically think about a product and the product is comprised of the app, the firmware, the hardware, the cloud, the probably the web browser version of your app, um, it all comes into play and it's equally as important to, um, you know, to dedicate enough care for each and every one of these aspects of your product. And I don't think one will exist without the other. So, um, again, that doesn't mean that you have to do it in-house. You know, we, that's where we step in, uh, with our expertise for many startups and scale ups.

But having a firm mono, like a helicopter view of your entire program, of your entire, um, value prop is, I think is really crucial for the success of theirof.

- [Przemyslaw] Yeah, I fully agree with Peter that, uh, the technology is, of course, important, but to them, at the beginning, you have to take care about the helicopter view everything around the technology to define what do you want to do.

- [Ryan] Yeah, no, I totally agree. Um, it's understanding what you're trying to accomplish. Um, sometimes, you know, thinking from almost the end user backwards and then really figuring out what all the pieces you need to put together, um, to bring a solution to life or to which solution to adopt, I think is super critical and not enough times do companies focus on that for sure.

- [Peter] And also I would even add that, you know, technology aside, because building, again, this is a well kept secret, but building is not the hardest thing out there, right? Uh, validating your concept and making sure that you're building the right thing that will resonate with the market that will find it's target audience, that will get a, you know, get the product market fit. I think that's the hardest thing. Once you figure this out, you know, what you want to build then building technology is relatively, um, straightforward.

- [Ryan] How do you kind of approach finding product market fit, or how do you kind of approach getting to that point where you're confident in, or at least you know, more confident in what you need to build? And now, like you said, building is probably the lesser of the harder parts. How do you kind of advise companies or how do you all think through getting to that point where you feel like, okay, now we know what we need to build and we feel like we're in a good spot?

- [Peter] We advise our partners to dedicate a lot of time to the market research phase, the validation phase before we even write a single line of code. We may do that for, you know, dirty throwaway prototypes that we then validate with actual users. But I would, uh, I advise everybody to, you know, and again, I'm an engineer, so I kind of, you know, um, uh, have to,

um, cool down my aspirations and my passions to, you know, to get started, to get building. But actually the wisest approach, in my opinion is delay building as much as you can until you get enough of a conviction that what you're building, what you're, what you've dreamed about is will actually find a, you know, a customer.

- [Ryan] Yeah, it's and does that change at any point in or any, with any of the components within a solution? I mean, obviously software and hardware are quite different, right? So does that approach change in your mind when you're approaching the hardware or approaching the software? Or is it kind of the same?

- [Peter] It's, of course, it's different because once you, you know, arrive at a piece of hardware, it's relatively hard to change it, right? Whereas in, you know, in software, you can keep iterating on, you know, on through means like over the air updates, right? So, of course, we advise our partners to, you know, spend extra time making sure that the hardware that they're building, that the build and materials is, you know, is exactly what they want to do, right?

So and has enough of a evidence in user research, user interviews, you know, other validation methods to, you know, to start manufacturing it.

- [Przemyslaw] Yeah, and hardware is something physical, so you have to not only design it by technology, but also the shape, the enclosure, the how this device have to look like. This is also important. So, we are trying to spend a lot of time to speak with the customer, to define what is possible, what we can do, what is possible right in relation to them, to the technology of our product producing something.

- [Ryan] Absolutely. Now, when it comes to, when we get to the point of like ready to deploy a solution, there obviously are more challenges that arise then and also after a deployment. You know, we talk about at times how many solutions get into the pilot stage and then fail. But when it comes to, just like at a high level, the deployment of an IoT solution, where do you all see most companies that either you are observing or maybe have worked with in some capacity or they've shared from past experiences, they haven't had maybe the greatest success with deployment.

So, you know, where are they failing? You know, what are some of those stories you may be able to share with us?

- [Peter] Okay. I'll think about specifics, and may throw in one or two examples in a bit. But, I would say that the most, the most underestimated thing is the amount of work you have to put in, into the product after launch, right? Everybody looks up at the deadline of, you know, when are we going to ship it?

Right? And that's this glorified moment when a lot of people actually think that the, you know, the tech team, the engineering team, will stop working and, you know, focus on something else. But the fact of the matter is the, once you deploy the product, once it finds itself in the hands of the users, it, the game really begins there, right?

Then the you don't have people who deeply understand the product, who use it. Now, the product ends up in the hands of people who, you know, perhaps have no idea. They are not as technical as, you know, we or all are. The amount of investment into customer support is minimal and in many cases, in many cases we worked on. So, I think making sure that the, you know, the users that are using your product are well taken care of is, I think is the single most underestimated aspect of, you know, post-launch, right? So making sure that the UX is there, that the product is obvious.

That, you know, people have some, you know, phone number to call in case you know something doesn't work. And returns are really expensive, they can kill your company.

- [Ryan] And I imagine like just being able to get feedback and learn from the interactions and the engagement that those end users are having with the actual solution contribute a lot to the ability to make changes, iterations, send updates. You, you know, just ensure that it's working as properly as possible.

Maybe catch some things that could lead to failure ahead of time.

- [Peter] Of course, of course. And you know, there are so many tools that, you know, you can employ both on the software side and hardware side. You know, telemetry, crashlytics, analytics, all these things will help you tune the product so it best serves users. But again, I think more often than not, people think of, you know, the ship date as this finish line, whereas, you know, we often perceive it as a, you know, as a starting point really.

- [Ryan] Yeah, absolutely. And since we have your Head of Hardware here, I should ask on the hardware side, do you kind of see any failures or issues kind of coming from that end of the world?

- [Przemyslaw] Technology is going very fast. So, we are changing the life almost day by day. Yes. Starting from the, I don't know, five years ago, we start with Bluetooth and now a lot of devices is connected over Wi-Fi and then other specific connection. Narrow band, ultra wide band to measure the distance between the devices and so on and so on.

So it's really, this is part of the work. It's really, really fast.

- [Ryan] Absolutely. Now what about, so let's think about, you know, we've talked through deploying, and you mentioned it's kind of more of a starting point as opposed to a finish line as other people think about it. But you know, as a solution is deployed, scaled, there's always going to be questions about, well, what happens when technology changes?

What happens when we install a new system internally? Or what happens when things need to be updated? How do you go about that future proofing of the hardware and software to, you know, the best of your ability? Obviously you only know so much, but what are the steps or what is the thought process like or advice to kind of share with our audience on how to think through future proofing both the hardware and software side?

- [Peter] Sure. So, you know, it really starts with the validation phase, right? So making sure that the product market fit that you'll find is then build the concept of what you're building is built on the right stack, right? So making sure that your firmware technology is, you know, is futureproof, will not go away.

Making sure that your mobile development stack is something that is relatively futureproof. Of course, there are so many tools out there right now, you know, starting from native development with React Native and Flutter and so many others that, so, you really have to think about, you know, starting with technology, making sure that, you know, you will be able then to find people who will drive this technology, you know, because it's, you know, it's not a sprint, it's a marathon. So, you, just making sure that the talent will be there in, you know, in three, four years, right? Because your product will be there. So making sure that the talent to, you know, to keep iterating, keep iterating on the technology side of things and the software side of things will be there.

Making sure that you're building your solution on the relatively future-proof backend or cloud services, right? Uh, we all know the story with, you know, with Google's IoT wrapping up, offering, wrapping up. So, and I, this is not necessarily some something that, you know, people could foresee, although, you know, Google's track record is what it is, you know, in terms of killing products really quickly.

But, you know, you really have to think through a lot of things starting with your software stack through the hardware components, making sure that these will be available down the line. And then, you know, making sure that you have means to, like we've said before, right, to listen and to understand how people are actually using your product.

So having, you know, closed look with analytics, with telemetry, uh, getting raw data out of, you know, your device's sensors. All these things come into play, right?

- [Przemyslaw] And it's also another thing related to the hardware, because I'm more hardware than software. For sure, you have to be ready for the future updates. So it doesn't mean that you have a methodology to update your device, but you need also some memory. You need some MIPS on the, your microcontroller to, for the future, future features let's say. This is also important.

- [Ryan] No, totally. Absolutely. I mean, it comes from both sides. I mean, we talked on the podcast before about network sunsetting and things like that. And how do you prepare or protect against that, which obviously you won't only can do so much, but it's with recent kind of network sunsets, it's become much more popular conversation on how to prepare with more devices in the field.

A lot of devices you may not be able to even physically interact with at times. So how do you prepare for that kind of thing? And that carries over to the hardware pieces, the software pieces, kind of everything, even outside of just the network. So it's definitely a question that a lot of companies I'm sure have when it comes to adopting.

As well as building and launching these solutions for people to adopt. Is there a way really to kind of ensure product success? I mean, there's just so many variables at times I feel like, and, but obviously the industry has come a long way. We're learning so much with every deployment, with every new technology, it's technology's getting more mature.

How do you all kind of like, you know, I guess how would you answer that question if someone was to say, can you ensure that my product is successful? I mean, obviously, we only can do so much so, and we, you know, you all can only do so much. Right? So how do you kind of think about that kind of maybe thought that I'm sure customers have?

- [Peter] Yeah, you said it, you, we can only do so much, right? Collectively. So, I don't think that there is a single way to ensure that your product will be successful. You can do everything right and then trip over something that you haven't foreseen or there can be a competitor that was, you know, in stealth mode and just popped up.

I would say coming back to what we've said before, invest as much time as possible into validating your concept against real users. I mean, I love to, you know, listen to my gut, but it's not always right. You know, I'm often surprised by, you know, what people actually need out of technology.

And these are, you know, rarely things that I expect from technology. So really invest a lot of time, push that, you know, that moment when you start writing lines of code, prototyping as far as you possibly can and make sure that, you know, you have enough of a market signal to, you know, to pull a trigger on an idea.

- [Ryan] Yeah, I agree. I think, I've had some guests recently talking about what you can do leading up to any, even any physical deployment with, you know, simulation, using digital twins, other types of technology to test as much as you can, which again, can only do so much. And then understanding the industry that you're building for through hopefully the customer you're working with and their expertise and domain knowledge. But again, that only, you only can get so close to thinking, okay, we're ready to go, ready to build. We know this is gonna be valuable. If you're selling especially into, you know, or you're building for somebody who is then selling to their own customers and hoping that the demand is there.

But when you build directly for one company, they still have to see an ROI, right? Right? And they can only show so much of an ROI through a pilot, but once they get it deployed and start to scale, that's when the real value is. So being able to show that win early on so that they can justify internally the cost and the time to scale, I think is, is a big challenge.

But and when you do that, it leads to a lot of success for companies across the board.

- [Peter] One hundred percent. And it's easier, it's easier than ever to, you know, to build stuff. Right? So it really boils down to, you know, to finding that niche, to finding that passionate group of users who will be your, you know, your power users.

- [Ryan] Totally agree. I guess this leads into a good question that I wanted to ask you before we wrap up here is, you know, we've talked all the way through where to start, how to deploy with a better chance of success. How to future proof, you know, those kinds of things. But let's say we get through all of that successfully, the pilots of success, we're ready to scale.

How do you go from 10 devices to a hundred to a thousand and beyond successfully? I'm sure a lot of that is, starts earlier on than you think as far as how to prepare yourself for it. But what is, what do you, what is, when we get to that scale phase, what's different about that or what do you, you know, what do you now need to start thinking?

- [Przemyslaw] Okay. You have to think about this just at the beginning because it's not so easy to move from, as you said, 10 pieces to one million or even more. If you have to change it in, in the, let's say in the meantime. So, at the beginning you have to start to think about, one, the production does it look, looks like, how to make the enclosure, how to make a test automatically. Use as much as possible, the automate, automation of this process and so on and so on.

- [Peter] Yeah, it's really, like I said before, it's really easy to start building, you know, there are so many ready to use building blocks. But if you want to scale, you know, beyond the initial 10 pieces that you know, you will quickly, you know, bang out in your home.

You really need to think about, you know, the component, what components you're using are, are you able to source these components? Where are you going logistically, uh, to the manufacturing? Who is going to be putting together, you know, the enclosures and, you know, distribution and logistics?

This is something that people rarely think about when they're, you know, they come up with their first initial prototype based on Arduino, which is again, really, really simple to get going. But then, you know, then all class of new problems start when you're, when you're thinking about, you know, producing a thousand, ten thousand, or a hundred thousand devices.

- [Ryan] Yeah, very good points. I mean, it's, it's great when you get to that stage to kind of start to scale, but obviously it brings in different challenges and thoughts that you need to really think through. And I, and I know it varies depending on the use case, the industry, the environment, what technologies you're using.

It's, it's a lot, but I think it's important for people out there to understand that if you find the right company to work with, to bring your kind of solution to life or to market, however you're approaching it, a lot of this has already been experienced, learned from, kind of adapted for. So it's important to understand that even though it can be overwhelming at times when it comes to thinking about how a solution's gonna be put together, people out there like you all know how to do this.

It's just can you, you know, can that, does that company have a real problem, a real business case? And then proving that out and then, you know, working together to make it successful. I think is super important to remember. Let me ask both of you before I let you go. So, you know, as we head into the rest of this year, what, what are you kind of most looking forward to, excited about from an industry perspective?

Is there anything that we should be keeping an eye out for, from each of your kind of lenses? Just out of curiosity.

- [Peter] I'm really astounded by, you know, the progress I see with large language models every day. You know, it seems like we're making a really ginormous leap almost every day on a daily basis. On a daily basis, I see new products that utilize those large language models, and I think it's just a matter of time when these spill over to, you know, IoT devices, wearables, personal assistants, and all the devices that, you know, we're really interested in building.

- [Przemyslaw] I see the same story with machine learning but related in correlation with the sensors with some wearables. Try to learn our body, our trainings, and then we can get some more information about, what we have to, I don't know, eat, how to make trainings and so on and so and so on.

- [Ryan] Absolutely. Yeah, it's a very exciting kind of space to follow for sure with a lot of different things going on that, I'm sure you all have your hands in pretty greatly. The language models is super interesting. I just watched a podcast yesterday kind of talking through it and I was actually having a conversation earlier today outside of the podcast just about how a lot of companies out there when it comes to implementing AI, in their solutions are more so kind of doing it for the marketing purposes to feel included, but they're not necessarily truly bringing AI capabilities into their solution. So, how are companies really going to implement powerful AI? Now, I know at the same time, I know that, you know, AI requires data and a lot of times, when it comes to just thinking about enterprise AI, you have, you know, IoT's gonna be a big driver of that data, which is why they go together so well.