Aaron Zarraga began his career as a hardware engineer at Amazon in San Jose on the Concept Engineer team.

Leaving his full-time job with Amazon to co-found a hardware startup (Sensel) was a difficult decision with lots of learning moments for Aaron.

We cover Aaron’s professional background, and Aaron shares hard-earned advice for engineers interested in the development process of hardware products:

Aaron Zarraga


Play in new window || Download


Max: Welcome all! Max of the Accidental Engineer here.

Today we are joined by Aaron Zarraga. Aaron is co-founder and CTO of Sensel. Do you mind sharing for our audience a little bit about what Sensel is and what you guys are doing?

Aaron: Sensel is developing, basically, the future of touch technology. We’ve developed a very high-resolution, full-force-sensitive touch technology that we want to revolutionize any consumer electronics, any human interface. We want to use it to build the touch screens of the future.

the Sensel USB device

It’s full pressure-sensitive, it works under water, it works with any object. It’s a very natural way to interact with computers and machines, and I have a demo that I can show later in the video so you can get a sense of what we’re making.

We also have a product on the market right now called the Sensel Morph, where we took this new technology that we developed, bundled it up into a really fun product that lets you take advantage of the new technology we built in many different use cases–art, music, gaming and productivity. It’s an interface that can morph into a bunch of different tools.

Max: Like Aaron said, as we get further in the interview we’ll do a live demo of Sensel’s product, and we’ll pull out my MacBook and the USB device and plug it in and do a demo.

For our audio-only listeners, we’ll also be narrating a little bit about what we’re doing, so you guys can get a sense, but check out the website to see screen shots and whatnot.

But for our audience that is used to our format of asking about a guest’s background–do you mind sharing a little bit about how you got to co-founding a business a few years back?

Aaron: I went to school at Stanford, studied electrical engineering there, did a B.S. and M.S. degree there. And then started working at Amazon, worked in Cupertino in their hardware company, Lab126.

We were in this really cool group called Concept Engineering. We were working on very future-looking products. A lot of the new product concepts that came out from Amazon were actually first developed in that group. I really enjoyed that team, and it was a really fun first out-of-school experience. While I was working there I met a few other people, and one of them turned out to be my current co-founder, Ilya Rosenberg. I met him while I was part of that team. And we had been doing a lot of work in the input space and human-machine interaction in that group.

Ilya left that team back in 2013 and started thinking about to build a new touch technology of the future, and he spent some time in his garage making some really interesting new concepts for how to build a touch sensor. And then he came and found me–contacted me one day–and I went over to his house and saw what he put together.

I knew that with he and I both working on it we could make it something really, really special, and I knew that it could have the ability to transform interfaces. That’s when I decided to join him. And we started Sensel in December 2013.

Max: How critical was your bachelor’s and master’s degrees towards preparing you for that first job at Amazon and then, ultimately, for preparing you for co-founding a business?

Aaron: I was an electrical engineering major—electrical engineering is one of those skills that’s really hard to learn on your own. I’ve met a lot of really talented programmers that can self-teach themselves computer science. For electrical engineering you can’t go out and easily get all the equipment that you need to do electrical engineering as it’s really expensive.

You can’t just do that on your own, and that’s why it was really important for me to do that at school, because they have the labs and they had all the classes where you can go and you can start using their equipment to build stuff. You learn how to use all the test equipment.

I think it was really important just to get me in the gear of how to build things and how to really engineer stuff. I think it was really important. And, in terms of running a business, right? Electrical engineering degree doesn’t teach you a whole lot about how to run a company.

For me, school is more about teaching you how to learn and teaching you how to pick up new things. I’ve used that to learn what it takes to run a company, what it takes to go raise money, hire employees. All that stuff is brand new.

I think, when you’re doing a startup, no matter what background you have, you’re going to be drinking from the fire hose the whole time. I don’t think you can prepare by taking classes in it and stuff. I took an “EE for Business” class, but nothing really prepares you for the stuff that you’re going to learn on the job doing startup.

Max: I know you’ve given back some to people by mentoring who are interested in doing similar to what you’ve done.

Aaron: Back in 2014 we went through StartX, which is Stanford’s startup accelerator program. They have this great crash course in “what are the things you need to know to get this startup off the ground?”

We learned how to raise money, basics of how to do a product launch, and things like that. I stay heavily involved with StartX. They have a hardware neighborhood, which is part of the program where all the hardware companies get together and talk, and I’ve run that session a couple times with another team member.

Sensel just turned four in December, just a week ago. We’ve been doing it for four years now, and I’ve made a lot of mistakes along the way, and we’ve done things the wrong way, and we’ve done some things the right way.

If there’s anything that I can share with someone who’s going through that same journey and help them, I love doing that. It’s something I really enjoy.

Max: People can feel free to get ahold of Aaron, we have a profile page for you where people can submit comments or questions. If you want to get ahold of Aaron, you can ping him there.

Aaron: Yep.

Max: One of the things that I’m personally curious about is, basically, that hardware products are a very expensive series of sunk costs-

Aaron: Yep.

Max: … to get them to market. Do you mind sharing what those look like from a very high level? Not necessarily specific to Sensel but–in contrast to maybe your friends who picked up programming without a formal degree and were able to set up a website or whatever–what are some of the big hurdles of going to market with a hardware product?

Aaron: I have a interesting … ‘cause my path was starting at Amazon, right? And I was an engineer at Amazon, and we’re cranking out hardware products, right? But what’s interesting is, you only really see the engineering design side of it, at least when you’re a design engineer. You see the iterations and the prototyping, and you start getting towards production. And that’s a lot of work. I got a glimpse of what it takes to ship a product from there. What was fascinating is, then, you go outside of Amazon or whatever big company, and you’re doing a startup and you want to do hardware, you realize how much stuff, infrastructure, is already done for you at a large company. I remember, being at Amazon, I was a young engineer just out of school. There’s a whole operations team, and there’s a whole marketing team, and I never interacted with them, being in a entry-level position there, and I always wondered what they did. And you learn very quickly, when you’re on your own in startup, you’re like oh, this huge thing that I have to do, that’s what the whole operations team was working on.

That’s what the marketing team was working on. And at Amazon, I would just show up and there’d be a factory, and I’d go to the factory, and you solve problems and you go home. And you’re thinking wow, that was really tough. At a startup, there is no factory, right? You start thinking about, we’re going to build this product. We’re going to do it this way, that way. And you’re like, now I got to go find a factory. I’ve got to go find a contract manufacturer. And then you’re like, well, how do I do that? And you start talking to experts, or we had a lot of help from consultants that would help fill in the gaps. But then you go and you’ve got to negotiate an agreement with the factory, and you’ve got to do payment terms. And each little bit like that along the way is another huge can of worms that you start opening. And one of the things that I’ve told the companies at the hardware neighborhood is that hardware can be great.

I mean, there’s nothing like building a physical product and shipping it to someone and having them use it and seeing your product out in the real world, but it is tough. It is a big undertaking. I’ve worked with some companies that were doing software stuff primarily and they wanted to build a hardware widget to be an accessory to the software. And I’ve talked some of them out of that, ‘cause it’s like, I think what you’re talking about is great, but unless you have the funds to go pay some company to design that whole thing and manufacture the whole thing for you, you should probably think twice about it. Because just building that little widget seems really simple now, but it’s a huge effort just getting that done.

Max: Oh, I believe it. Do you mind sharing, for our audience, a little bit about, perhaps, a factory negotiation that you’ve had to be involved in?

Aaron: Well, think about a contract manufacturer (CM). I knew that there was factories in China, factories in the U.S., but I never knew that there are different types of cm’s that specialize in different things. There’s some that do medical equipment. There’s some that do consumer electronics. There’s some that only do high-value, low-volume products, and then there’s other factories that are set up to only do high-volume, low-margin type things. That was a huge learning experience. Just going in and figuring out what type of CM makes sense for your product, where you can try to hit the cost targets, then you’ve got to down-select and make sure that cm wants to work with you. This is the interesting part, is that you’re founding a company, you’re building this product, and through your eyes it’s like this product is a slam dunk, right? Everyone is going to want to build this. Not necessarily true. You go into a cm and they say this is a cool product. You have to pitch them, right? You pitch them.

Even though you’re going to be paying them, you have to pitch them that they should work with you. And a lot of them have aversions to working with startups now, because I’m sure they’ve been burned in the past. And then they’re asking, what kind of volumes? It’s like well, I want to start by building a few thousand or something like that.

Max: That’s not enough!

Aaron: “We’d rather start it … come talk to us when you can build a million or hundred thousand.” And you’ve got to find the right size CM, and you’ve got to work your way up. And then, the agreement stuff is all after that. You’ve got to negotiate terms, etc. But yes. That’s one very small part of the hardware stuff. But each one of those is … stacks up, becomes big.

Max: Is there is a class of contract manufacturers that’ll say yes to anything despite their quality guarantees, that you had to be wary of?

Aaron: Yeah. It’s one of those things where people say “if it sounds too good to be true, probably is.” There are a lot of companies out there that will say yes and tell you anything. And then, you actually find out ooh, actually the company may be going under, and they’re just trying to get a last-ditch effort and-

Max: Collect a check?

Aaron: Yeah. And you’ve got to do due diligence. Even for a CM, we asked for references. We said, you’re pitching us that you’re set up to work with startups and you can work with us to ramp us to high volume and stuff like that. Let me talk to someone you’ve actually done that with. And if you can’t find someone, if they can’t point to one person that they’ve gone through that process with and been successful, then you should be a little wary.

Max: And to cope with this problem of negotiating with contract manufacturers who say you don’t have enough units, are there any organizations or groups of hardware startups that group together to negotiate as a group with contract manufacturers?

Aaron: I don’t think so. And if there is one out there, I’d love to know about it. I mean, I think that would be a fascinating model where you have a group of startups that approach cm’s as a group. But it’s tough, though, ‘cause even in that situation where you’ve got 10 startups that each want to do 2000 units, now you’re talking about a bigger number, each product is so different. And they have to do so much work just to build the manufacturing line out for each product. And that’s why it’s tough. And it’s like, from the factory’s perspective, there’s a reason that they want that volume, because they’re going to put in a bunch of fixed cost just to get the line up. And then, you’re telling me we’re going to spend six months building all this production line and we’re only going to build 100 units? No.

Max: FairMany of our audience, engineers or not engineers, lead a pretty chill life where they go to a day job at an office in the San Francisco Bay area or other large U.S. city generally.

Do you mind painting a picture for people what a day for you might look like?

Aaron: Well, to give you an idea, today’s Saturday and I’m here. It’s great that I’m not in the office right now. I think that, as a founder of a company and even working at a small startup, I’m sure a lot of the people out there know this, but you don’t have the luxury of assuming that your work is going to be there tomorrow. Right? At a large company, I remember the mentality was always hey, it’s 8:00, it’s 7:00, just going to go home. The work will be here tomorrow, and don’t worry about it. And fortunately, you can be in situations where you don’t have to take your work home, you can turn off your email until the next day, you can take the full weekend off. That’s really great, and I really do think that if … when you’re in a part of your career where you’re in that, you should really appreciate it because not everyone gets to do that. And I look at back my time at Amazon, I was like man, I had so much free time working at Amazon. ‘Cause then I worked long hours, but I still didn’t bring my work home on the weekends. But at a startup, you don’t have that luxury. It’s like actually, if I don’t answer this email to this investor right now, the work might not be there tomorrow. You know?

It’s those type of things you’re dealing with. Everything is very important. Everything is urgent. Everything is very critical. And you’ve got to, sometimes, turn on stuff on the weekends. You’ve got to stay late.

Max: Does your work take you overseas at all to talk with contract manufacturers or otherwise?

Aaron: Yeah. That’s actually a really fun part of the job. I’ve been to several different places abroad, Taiwan, places in China. I was recently in Japan visiting manufacturers and suppliers and factories, but also potential customers as well.

Because while we’re developing our own product, we also have this core technology that I spoke a little bit earlier about, that we’re also working with customers to integrate into laptops and working with smart phone manufacturers. All those guys have a large presence in Asia. I think that’s been really great from a learning perspective, just learning. It’s learning how the business customs work in Japan versus Korea versus China. It’s all so different. That’s been a huge learning experience for me, and I feel very fortunate to get that opportunity to travel.

Max: When you do go on these travels, this might be a personal question, this might not be super relevant, but when you travel to Korea or Taiwan, what does a business trip look like for a hardware developer like yourself?

Aaron: It depends. Actually, we have a few team members that are traveling right now, and a lot of their trips are very different. One person is traveling to deliver some custom sensors he developed to a customer. He’s going to be working with them, in their factory, to integrate the sensor, and things like that. We’ve got a couple other team members that are in a different part of Asia that are visiting suppliers, because we have a network of suppliers that all provide parts for us. And when you’re building a new design or you’re building a new anything for the first time, you really need someone on the ground there. They’re there supporting the first-article inspection of these new supplier parts. And those two experiences are going to look very different. In one case, you’re the customer, right? And that’s one dynamic. And in the other case, you’re working with the customer. Totally different dynamic. And I’ve had situations where I’ve had more low-key trips where I’ll have a few meetings and I’ll maybe get some time in the evenings to go see parts of the city I’m in.

I’ve had other trips where I show up at 8:00 a.m. at the airport totally jet lagged, I go to meetings all day, I spend the whole week in meetings, and then leave, and then I only see the office there that I’m visiting. It really depends, but …

Max: Fair enough.

Aaron: You have really good ones and fun ones, and then sometimes you have ones that are just you’re just grinding all week.

Max: We had a guest on where we talked a fair amount about quality assurance (QA) and testing, and this is more relevant than ever in hardware engineering–especially, where you have statistical benchmarks maybe, or you have service-level agreements with partners who are providing you guys supplies or end products.

Can you speak to, for our audience that are not involved in hardware engineering—share a little bit about what QA looks like in hardware engineering?

Aaron: Well, you have the first level, which is just the specs, right? One thing we’ve learned, time and time again, is that … one of our team members always says, “Garbage in, garbage out.” If you send a spec over to a supplier and you say, “Here’s what I want,” and it’s garbage, and if it doesn’t clearly define everything that you need and clearly communicate everything you’re expecting, you won’t get what you’re expecting. You’re going to get the garbage out. It’s a transition. When you’re a startup and you only have a few people, you only have so much time to put together the spec, but we’re going through this process, we’re growing up as a company and getting much better at really knowing what our specs look like for each different component. We know we have our suppliers we’ve worked with to make sure they can meet the specs. That’s a big part of it, just the upfront communication. I want you to build this, and I need to go make sure that you know how to build it, you understand what I’m asking for. And then, there’s the whole part of QA’ing after the fact. If you’re the supplier and we already did the spec and you say, “Okay, I can build it,” then I need to go make sure that you’re testing correctly, that you’re testing the first line of defense, right?

Don’t let bad parts leave your factory. And then, I’ve got to set up tests on my end, so that when I receive them I can test them right away to make sure that they’re good parts. That’s, again, one of these things that’s this huge effort that you don’t really see. You see a little bit of it when you work at the big companies. You see there’s people on the team that are QA, but you don’t see, on the supplier end. I mean, you just see parts getting to the factory, a lot of times. But on supplier end, there might be a whole set of testing fixtures and things like that, that someone at your company’s developed to ensure that you’re getting good parts out. It’s definitely a learning process. And you learn who your suppliers are. Some of our suppliers we know, this supplier’s never shipped us anything bad, and we know that the person in charge is running the show, that they know what they’re doing. Others, it’s always hit or miss. We’ve got to really double down to make sure QA’s good and we’re there in the factory.

Max: The parallels with QA and getting the specification right upfront in software engineering are totally direct.

In running an international business, when it comes to writing specifications, is there a language gap? How do you cope with a language gap dealing with contract manufacturers whose first language might not be English?

Aaron: It happens a lot. With suppliers, and even some of our customers we’re working with, it happens a lot, where you’re not really sure that you’re talking about the same thing.

A lot of it is just being there. That’s one of the things that I’ve been … I was hesitant, really early on, about “hey, we shouldn’t be traveling. We can do everything over videoconference.” But it’s amazing—even over videoconference, you can only get so far. If you have just two engineers in the room, even if they’re totally not on the same page language wise, and you can just have them interact in the same room and you can whiteboard stuff out, and you can show different things, different concepts you’re thinking about, it’s just so valuable. And I think that’s one of the biggest things that we’ve learned, is that you need someone there. I think we’re looking at establishing a team in Asia so that we can have on-the-ground presence without having to travel there. I think that would be really, really great. That’s a whole ‘nother challenge that we haven’t gotten into yet.

Max: Well, this is a good opportunity to plug something that I’d love to plug, which is whether there are any roles that Sensel is hiring for right now that we can advertise?

Aaron: Oh, yeah. Software engineers, we’re always looking for software people–especially software people that want to work with hardware stuff.

If you have any interest in working with a hardware peripheral or doing any firmware development or any kind of factory test procedures or … we’ve got crazy test robots that are moving around, doing all this stuff.

If any of that sounds interesting, definitely let me know. Always looking for software. And then, I think we’re going to be growing pretty soon across the board, so any mechanical materials engineers, even operations. If you’re into hardware and you’re interested, just shoot me an email, let me know, and love to chat.

Max: So far we’ve been pretty abstracted at exactly what Sensel’s product is. We’ve talked about it being a peripheral to device, whether smartphone or laptop or what have you, but I figure now would be a good time to do the demo?

Aaron: Oh, yeah. That sounds good.

Max: I’ll bring out my MacBook, and Aaron will bring out Sensel’s device.

Aaron: I just want to show the device here. This is the Sensel Morph. This is basically a large, tablet-size touch sensor that features our new technology. You’ll see when I start the demo, but our technology senses not only where you’re touching, but how hard, for each contact. It’s almost like a camera for force.

Current touch solutions today, if you touch your phone, it only knows where, right? You only get X, Y coordinates for each finger. But we’ve got full pressure. We also support any objects. You can interact with paintbrushes or any kind of interesting materials. This is our product, the Sensel Morph. The idea is that we have a base sensor. We have an API so you can get all the data out, and things like that. But we also wanted to make the technology available to even non-engineers. You don’t have to be a developer to use it. We’ve developed an ecosystem of, basically, physical applications. We’ve got different interfaces that snap onto the device.

Max: For example, a keyboard?

Aaron: Yes, this is a QWERTY keyboard with little silicone clickey keys. And basically, when you put the overlay on, the whole device changes its behavior. They’ll just output events just like a keyboard. And then, we’ve got a piano in here. You snap it on, and then it’s like a full MIDI controller with pressure. We’re also [inaudible 00:26:50] MPE, so it’s full pressure sensitivity, and things like that. The basic idea is that we have our touch sensor here. When you plug it in your computer, you can fire up a visualizer, and you can see that you can tell not only where I’m touching, but also how hard. It’s full pressure sensitive. And it’s really, you can fit as many points as you want on there. If you see, it’s seeing a full-force image of my hand. And we first developed this product, we took it to Hackathons, and we had all sorts of interesting, really cool applications where people were using it as a drone controller and using the pressure to fly things around using, really, the three dimensions of the sensor for controlling things in real life. We’ve had art applications, music, even gaming, IOT using it as a gestural smart surface and things like that. And then, I mentioned that we also sell a variety of different overlays. I just have a couple of them with me today. This is a piano, so this is a music application.

This is our keyboard QWERTY application. And the idea is that these are just flexible overlays. When you put them on our sensor, we can actually sense the touches right through the materials. Because we’re sensing force, we can sense through these different flexible overlays. You can see here, it’s detecting not only which key I’m hitting, but also how hard. We sense things like … you can do pitch bending and modulation, and things like that. It’s a very expressive and fun device. We’ve made it so that your average consumer that doesn’t want to program can just throw this on and then it becomes a MIDI controller, or you can throw this on and it becomes a keyboard, so you can switch views and just start typing and do whatever you want, hello world kind of thing.

And it’s got full Bluetooth and batteries also. It attaches to tablets and iPads and stuff like that.

Max: That was an awesome demo, for one–

One of the topics that I think would be really cool to cover is how, as part of going to market with a hardware business, the final step of getting the products into people’s hands. What are the big retail electronics outlets out there today for people who aren’t familiar with how people, very generally in the market, actually buy retail electronics?

Aaron: It’s interesting. Actually, it’s another big challenge getting into retail, right? Right now, we’re selling primarily through our site at Sensel.com. We also have a store through Reverb.com selling Morphs through there. We’re also looking at some other experiences.

There’s a really cool startup called Beta. They have stores that you can get your product in physical outlets across the country, and people can go try it. We’re looking at places like that where we can get the device into people’s hands without doing the full big-box retail like Best Buy and things like that.

It’s very difficult to get your product in there because, one is the margin. They take a big cut of your profits. You’ve got to make sure the cost structure looks right, and then oftentimes you have to commit a certain amount of inventory. It’s a financial risk and commitment that you really have to work towards.

Max: This is all information that I think–by and large–our audience is not familiar with, and it’s super illuminating.

Knowing what you know now, would you do it again?

Aaron: That’s what I think about a lot. I mean, it’s tough. I think one of the things that’s challenging is that startups, for whatever reason, have this perception that it’s almost glamorous–working at a startup and that it’s this dreamy situation to be doing.

And the reality is, it’s a grind. Every day, it’s so difficult and so hard. And I don’t want to sugarcoat that, because that’s how it is. You’re building everything from scratch. But at the same time, while it’s one of the hardest things I’ve ever done.

I mean, it’s definitely the hardest thing I’ve ever done in my career, for sure. It’s also one of the best things I’ve ever done, because I look at the amount that I’ve learned in four years. Four years ago, Aaron knew nothing about how to go raise money or how to hire an employee or how to do any of this stuff that you’re doing. And while it’s so tough and it’s such a demand ‘cause it’s not only the time, it’s the mental effort, the emotional effort of being invested in your company. It takes a big toll, but you’re just learning the entire time. And even when times are hard, you’re still learning that whole way.

And I think, “woull I do it again?” I think I’m going to stay at Sensel and push it, and I really think that we’re on a great path to grow an IPO. Post Sensel, whatever that might be, if I decide to … hopefully, it’ll be IPO and everything’s good, and then I can see what opportunities exist after that.

But would I do another startup? Probably not right away. I think I’d love to have some … a break from the startup space for a little while, but I think I would get into it. ‘Cause I think once you get a taste of seeing all parts of the business, I have a feeling that I’m going to continue to want that. Because my day to day, people always ask me, “What does your day to day look like?” And I can never give a straight response, because it’s like, “Well, this past week I was doing all this HR stuff and setting up all these different things, but this week I’m doing a bunch of engineering and working with customers. Next week, I’ll be doing something else.” Getting to touch all the parts of the business is just such an interesting and unique experience that I do highly recommend it for people that want to do it. You got to know what you’re getting into.

You can’t do it halfway. I think you got to commit. You got to, they say, jump off the cliff or whatever, and you got to commit and go all in. But I think once you do, I think it’s an amazing experience. And all the team members that we have at Sensel, we have 15 right now, and all our team members are amazing. And I’m like, it’s just so great that we … this is a team that we built. And that’s something we’re really proud of, and I want to work with all those guys throughout my entire career. It goes beyond just this current company. That’s real, and I really do think that’s been an amazing experience.

Max: Awesome. Well, we’re proud of you! Thanks for coming on, Aaron.

Aaron: Thank you so much.