May 15th, 2018
Making and checking parts is the most frustrating aspect of PCB design. Find out how Zak Homuth set out to change that and what is next for Upverter, EE Concierge and the future of PCB design in a browser-based setting.
- Why the Upverter parts library? Verified parts at scale, in a high-quality way.
- Verified parts - free to use for everybody coming soon!
- Verifying Datasheets, it’s a lot of work and at the heart of design frustrations.
- I wanted to take the magic of Github and Google Docs and create something for hardware designers.
- From concept to manufacturing in 20 hours.
- This is a conduit for bringing ideas to life.
Links and Resources:
Verified Parts on Upverter
A note about Verified Parts coming soon to Octopart
Hey everybody, it's Judy Warner with Altium's OnTrack Podcast. Welcome back, we have
another amazing guest for you today but before we get started, please follow me on LinkedIn. I post a lot of things for engineers and PCB designers and I'd love to connect with you, and on Twitter I'm @AltiumJudy and Altium is also on Facebook, Twitter, and LinkedIn. So please
give us a follow.
Well today I have a really extra special treat for you.
Zak, I think I'm gonna destroy your last name so I can say Zak?
Your last name is?
Homuth - see I would have killed it. So, this is Zak Homuth, he is one of the - are you a co-founder of Upverter correct?
So Altium acquired Upverter and EE Concierge in August of 2017. So a lot of people have asked me why Upverter, what was Altium's interest in Upverter, because it seems like sort
of out of our lane. So I thought I'd bring Zak in today and let you hear straight from the horse's mouth. So start out by - if you would Zak - just telling us about what is Upverter, and what is EE Concierge, and what you were trying to create when you launched that company?
Yeah sure, so Upverter is cloud-based, schematic capture and PCB layout. And what cloud-based means is, it runs in the web browser. You type Upverter.com into Google Chrome, it shows up and you can do your schematics, your layouts right there you know, order the
boards for manufacturing right from your web browser. You don't need to download anything, it's collaborative, which was kind of our big superpower for a really long time. Which is that a bunch of users can work on the same schematic, in the same layout, at the same time, in a very Google Docs kind of way. That's Upverter, and then EE Concierge came out of an experiment we ran back in 2015, trying to figure out what was the most frustrating part of doing a PCB design. And it turned out it was making and checking parts. And Upverter - until that point - had a shared global parts library and what that meant was every time I added a part, those parts were available for the community but without somebody checking those parts…
Yeah, that could be a nightmare.
-Yeah it was - it was really scary for people, they were spending a lot of time checking parts, they were spending a lot of time making their own copy of a part that was already in the library, a lot of that you know, wasted duplicates...
You can not have a messy library.
-Yeah and so EE Concierge grew out of that. It was our attempt to clean up and verify and guarantee the quality of the parts inside of Upverter, and so we built a small army of Electrical
Engineers all over the world. They work in an uber-like model where they can kind of log on, make a couple of parts, check a couple of parts, log off again. They can do it full-time, they can do it twenty hours a day, they can do it an hour a day you know, or an hour a week if that's what they want to do. And we built a machine intelligence to check all the work that they were doing to make sure that we had the best possible parts in the world inside the Upverter parts library and then, EE Concierge was born out of realizing that that's probably not specific to Upverter - probably every engineer has the problem.
Yeah there's that.
-so we built a plug-in for Altium Designer so that Altium users could leverage these verified parts which inevitably caught the eye of Altium and one thing led to another but - that you know that's EE Concierge. It was this idea of verified parts at scale, and then kind of outsourcing at scale, for electrical engineering. But in like a high quality way.
Right so, how many parts are in the library now?
Yeah so there's about a million-and-a-half parts...
-in the Upverter library of those million and a half, about 275,000 are verified these days. And to
different levels of verified, some of them just have verified symbols, some have verified footprints. It depends on kind of what era of EE Concierge they were made in. But it but we've got about 25,000 like really, really rock-solid ones and they're the 25,000 that most people use.
That's awesome. So what's the plan kind of going forward, to get the rest of them? You said you have some kind of machine learning to help you verify that. Like what's the path going forward to get the rest or - you know million parts like you said - not everybody's using a million parts right? There's infrequent wonky ones in there?
Yeah there's kind of two answers to that. So how we make this maximally useful to the most engineers is similar to what Altium and Octopart did, after Octopart was acquired we worked really, really hard to make sure that Octopart was for everybody. It wasn't just you know
Octopart for Altium, it wasn't just parts for Altium, it was Octopart for everybody, or CAD, you know Mentor, everybody. So we're trying to do the same thing with the EE Concierge, we
want verified parts for everyone. So in the next couple weeks we're gonna launch verified parts on Octopart, so we're taking that 250,000 parts and we're putting them on Octopart free for everybody to use. And you can download them in EAGLE, Altium, Upverter. We're working on Mentor, I think we've got Cadence or CAD as part of that, so like in any format, free. Just find them, download them, use them. That's what we want so that's kind of one avenue for the EE Concierge. And then the other is - this is a little bit more kind of futurist and out there answer -
But if you were gonna build an AI that could read data sheets, first thing you would do is have a
huge number of people read data sheets and enter that information very reliably into a piece of software, so that you can check it all. And so the kind of like dot, dot, dot - is we think if we get good enough at doing EE Concierge, and we do it for long enough, potentially we can read data
sheets with a computer, kind of our self-driving car version of the Uber model.
Like and then part of me goes; and you're assuming the data sheets are correct?
Yes it's a real problem and so we had to do a lot of stuff at EE Concierge to catch like, if TI ships a datasheet, and the datasheet has a problem, they'll rev the datasheet and then they'll rev the datasheet, and then not only that, but they'll reuse packages and they'll reuse symbols and the reuse bits and pieces of that datasheet across other parts that they make. We had to build a ton of stuff to be able to catch when they made an update to one datasheet and apply those changes to all the other parts to use the same bits and pieces of the datasheet. It's an ongoing thing - it's a hard problem for us but yeah it's a real - it's a real issue.
Well it's like amazing to me that we're here like...
-at all [laughter]
I mean just from being in like - I started in the industry way back in the 80s - and like everything was done by hand and all that, so just that we could possibly even get to that point - it's just so sci-fi to me - but it's amazing, it's so great. So tell us a little bit about how did you get it, tell us about your personal history and how you went down the Upverter rabbit hole and popped up here.
Yeah sure yes, so I was kind of into software and computers before that was a cool thing I was on the internet pretty early, I was I think five or six when my dad brought home you know our first computer and you know I tore it apart and tried to figure all that kind of stuff out. I was building video games when I was seven, and...
-of course you were [laughter]
-all that but I studied Computer Engineering at the University of Waterloo in Canada and Computer Engineering is a little bit of Software Engineering, a little bit of Electrical Engineering, and then you know all the physics, and boring, normal engineering stuff that you have to do [laughter]
Waterloo is a co-op University and so every four months you'd go to school, and then every four months you'd go get a job, and so I got a bunch of these cool jobs, kind of all over the world, I used it as an excuse to travel. And so I worked in Canada's capital Ottawa, for the insurance company that insures the majority of Canadian doctors, and so they had some really interesting data. And I built a search engine and a database for them to to be able to search when wrong-side surgeries happen right, you know when they operate on the wrong lung or whatever. I built a search engine so that doctors could figure out the root cause of some of this error because that was important for the insurance company to try to prevent this from happening. So I did that, I worked in Germany for IBM, I worked in India for Infosys, and then I got a job in Waterloo working for a company called Sandvine and Sandvine build what are called deep packet inspection - telecommunications service. And so ISPs would install these in their network and it would sit between their subscribers - the people who use the Internet and the rest of the internet - and it would look at all the traffic that flowed through the box to try to figure out how much is Skype, and how much is YouTube, and how much is pornography and Facebook and everything else. Because if you think about it, ISPs are kind of like the water utility. They know that they sold so much water but they don't really know what the water was used for, how much is watering lawns right, and so we were giving ISPs that kind of intelligence. Anyways, I started off as a lab tech, I tore the boxes apart, tried to figure out why they weren't working, put them back together and made them work and I kind of worked my way up to actually designing the box. So I designed two of them for Sandvine before quitting, and I quit because I was really frustrated that on our side of the cubicle wall there were ten of us working on the hardware that was so essential for this company to exist - but on the other side of the wall there were 300 software guys who had Git, and they had version control, and they had collaborative tools, and they could test their code by pushing a button on their computer. I had to carry a 120 pound server around a building, and use screwdrivers and shit to like - and that's, that's part of hardware - I never wanted to take that away from hardware but it felt like it could be easier and you know. We were using Mentor Graphics’ tools and I was frustrated by the archaic kind of 80s feel of it all you know?
I used a Mac at home - I couldn't use the software on my own computer if I wanted to, and I was trying to build stuff at home and this was kind of before IoT was like a thing. This is right before Arduino, before all that stuff. But I wanted to do that stuff at home and it was just so hard to do any of that you know from my Mac, from home, without a huge budget - it's an endless amount of time and at the time we were seeing cloud tools like Gmail and Google Docs and Github, kind of emerging and so I left, because I was pretty frustrated that if this is like the state-of-the-art, if like one of the most complicated telecommunications servers that anybody had built today, was built by some punk kid in this office in Waterloo, working mostly alone, using tools that felt kind of clunky and out of date, like there has to be something better than there and and so we left to build it and it was really no more complicated than: can we take some of the magic of Github, and some of the magic of Google Docs, and build a tool like that for hardware engineers and maybe it takes forever to disrupt Mentor, or Cadence, or Altium, or any of the big guys that have been at this for decades. But we assumed there were enough people like us that just wanted to spend two hours on a Saturday afternoon designing a piece of hardware. There must be something we can build for them - a little bit like Google Docs right, it doesn't have all the features, you can't make fancy tables, you can't do it all, but it's slowly taking over the world and we wanted to do the same thing for hardware.
Well I think you point out something that I really noticed. It seems like there's a block of people that are like my age, the old 80s people and we've kinda just built upon old... and then there's the next generation that were five, okay I was 20 when I got my first computer, you guys kind of grew up with these things in your hands, and it I think it's kind of hardwired in your brain. So I think, there's more efficient, better ways to do things and we are building on old Legacy stuff, so sometimes we just can't see it. And so I think it's really very exciting because I think people are gonna be: oh thank god, somebody's built something modern you know, on the cloud that thinks, and operates you know, which I think was the big draw actually for Altium, which we can talk about that a little bit later so - so the Upverter I was going to ask you, but you've
answered it partially, is why Upverter? Out of the various other things you could do, why didn't you go into - I don't know - you have a very entrepreneurial spirit, so there's lots of things you could have done. Like why did you pick this one thing?
Yeah, so when we started Upverter, I quit my job before I knew what I was going to do. I knew I wanted to build something, I knew I wanted to start a company, I knew I was kind of done with working for the man, and I recruited two of my kind of college roommates. So these are guys that went to Waterloo with me. We lived in this terrible, decrepit, run-down house next to the campus, near the engineering buildings. You know we lived together, we worked together, we did our co-op jobs together we - you know we were thick as thieves. But I called them up and I said you know, how do you feel about quitting your job and like doing something new? And they both quit like the next day and so we got together in this old decrepit student townhouse and we wrote down hundred ideas of things that we were excited about, things that we were passionate about, things that we believed needed to be fixed. I was shouting loudly in the corner that we needed to build this - this hardware tool that was the Google Docs for hardware. That was my passion, that was what I was excited about. But Steve and Mike, they had some cool ideas of their own. There was a bunch of stuff that they wanted to build that a lot of it is actually been built, and a lot of it ended up being quite substantially large companies. So our second pick our kind of the front runners....
So what was on your cut list?
-Yeah so the second pick was, we wanted to build drones, and this was before drones were
cool. We wanted to build very large-scale drones that would be towed behind container ships and provide a bigger radar footprint than the ship can have itself because it's so close to the
water and you'd do this for a bunch of reasons. But the really burning reason at the time was Somalian pirates. If you could fly one of these drones above a container ship, you could get ten or a hundred times the radar footprints so you could really move the ship before anything bad happened. If you wanted to. Anyways, a company ended up doing this, and started right around the same time that we did and ended up being acquired for something on the order of two or three billion dollars and so we missed that one a little bit. But but we just - we just didn't even know where to start on it. But it was - it was the second pick.
So we didn't really talk about this ahead of time, but you know I think I know around the time AltiumLive went down, there were people, kind of gurus in the industry, and they're like so -
ultimately it was in October, so and Altium acquired you guys in August and I remember some people, like it was like all the buzz, like what are you guys doing? Why that customer? And
so what I want to talk about is - who uses Upverter? I'm thinking makers, hackers, hobbyists and maybe EEs that want to be startups or do personal projects - like who do you think the Upverter audience is, and how's Upverter going to serve them, and how many people are on Upverter? Tell us about your ecosystem a little bit?
Yeah so it's the 'misfits' mostly, these are the guys...
The land of the misfit toys! I like these guys! It really is, these are the guys that are unserved by the eCad industry at large, they're using operating systems or tool sets, or computers that can't run traditional eCAD, they're in funny parts of the world, they're students, they don't have electrical engineering degrees. In lots of cases they're the weekend warriors that can't steal a copy of the eCad that they use at work and bring it home, they're makers and hobbyists and hackers yes, but they're also - you know we helped some Nigerian kids put their country's first satellite into orbit...
-and they couldn't have done it using traditional eCad tools.
That's cool I love that!
-Yeah we - some of the first augmented reality startups were built using Upverter. Like kind of odd stuff like that where you couldn't necessarily use a traditional eCAD tool, you couldn't necessarily iterate at the cycles that one of those tools would let you you iterate at. But also like children on the internet, and you know Mac users and all that kind of stuff where you just can't use a traditional eCAD tool.
But misfits mostly, we think of them as kind of an - not really the next generation of electrical engineers - but but very much a different breed, a different type of 'doer of electronics'. Well it is kind of grass roots though I think that we are gonna see more amazing things, like drones being dragged behind boats, that are gonna come up organically and kind of like you, I think you're the perfect sort of head of this brand, like 'yeah dropped out of school, this wasn't working, I don't like it whatever, so yeah I can see that happening over the next five, ten, twenty years. I think we're gonna see amazing stuff out of that space. So how many people, I don't know how you quantify that, have used Upverter, or use Upverter or actively log on a month?
So how do you do that?
At the time we were acquired by Altium, it was a little over 50,000 people used Upverter,
they use it in a very 'bursty' way, they'll show up - they'll work frantically for two days - ten days a month, and then they'll disappear and we won't see them for a long time ,and then they'll come back. And we correlate that with their kind of idea cycle right. They'll have an idea, looking very excited about the idea, they'll work on the idea, they'll do that thing, and then they'll go away you know, probably because they built the thing and they want to play with it. It could just be that their focus has moved and they're they're onto something new, and then we'll see them come back, six months a year you know, a couple of days later - depends on -how much time and energy they put into their ideas. But that's okay for us, like we we never aspired to be the daily tool like somebody like Altium is, like we aspired to be this conduit for people to bring their ideas to life and you can only be as useful as people have ideas right. So if you have an idea every day, we can be useful every day.
Most people don't - most people have inspiration quarterly, or a couple times a year and that's
that's what that looks like. We have many thousands of monthly active users so thousands and thousands of people log on every month, to work on their ideas, and their little projects. And the average project is worked on for quite a small amount of time, relative to you know, what you would expect from other eCAD tools. We'll see products going from conception to manufacturing in like 20 hours or less. And so that's pretty amazing, if you consider that's two or three days of work.
That's unbelievable, so if - you said something earlier that I wanted to ask you about and that was - you mentioned that people can go into manufacturing. Do you have sources like fabrication, assembly sources that are related to Upverter?
Yeah, and we've had this in a couple of different forms over the years. We've had, what we like to call the print button, kind of refactored a couple of times inside of Upverter. We're currently refactoring it again right now, and part of that is as a result of the acquisition. We have another
company we acquired, Siva, who does a lot of stuff in the manufacturing space, and we're refactoring our print button to use some of their technology and would be better linked up with Octopart, so it should be a better experience for our users as a result of doing it. But yeah,
over the years, we've had a button, you click it, you give us your credit card number, and a couple days later something shows up in the mail which, which is what you designed.
-And we'll do that again in the very near future.
And it - was that assembled also, or just the bare board?
-We used to just do bare boards, and then we experimented with assembly for a little while, the new - the newest, latest and greatest version - that we're working on right now will be fully assembled and it'll probably include whatever enclosure your device fits inside.
Oh my gosh, I mean I'm like ridiculously excited about this.
Yeah, it's gonna be pretty cool.
Okay so I always ask - I don't always ask this - but I wanna ask this now. Okay, are you a geek or a nerd?
I - - geek, but I don't know why. I don't know what the difference is really...
It's just your gut - open question.
-I think you're a geek cuz I don't think there's - I don't think - I think... whatever [laughter]
It's something we ponder here on the OnTrack Podcast. What is a nerd and what is a geek, we've...
-big questions [laughter]
We've decided what geeks are cooler; nerds seem more like, at least to me, physics - like...
-yes more deep on the science side but this has not been proven, so if anyone wants to comment below and tell us what you think a geek and a nerd is, we're all ears. So but you were geeking out there, this is why I stopped you.
-cuz I'm like, oh you are like totally geeking out and I'm tracking with you man, I'm like, oh this
is like - we're having a geeking out moment right now [laughter].
So I was talking to our Head of Operations the other day, and I was just saying that during that AltiumLive, people were asking why would Altium, a professional e-tool, pick up Upverter,
and basically I think Altium has a vision to kind of embrace every level of PCB designer and also embrace and serve those 'misfits', those marginalized, or that don't have access you know, those that can pay for Altium Designer and they do it professionally, well great...
But that we want to serve the wider community. So I was asking Ted Pawela about it, and he was saying - and I just wanted to get you to chime in here - is that sort of what we were
talking about earlier - is that cloud - I mean a lot of software programs are going to cloud-based.
You know there used to be security issues, but they've tightened those up - so technology is
moving towards cloud based, and if we don't sort of pay attention and go that way too, I think will be sort of left behind and that, also the next generation, or the new, or the upcoming, or the
grassroots, organically-grown innovators, I think are gonna - like you - are going to be cloud natives right, and are going to be frustrated, like you were when you were at Waterloo. And to
also meet those people where they are, not expect them to cough up the money or fit into our model, but figure out where they're going, and what their model is, and what their needs are. And so that was - that's what I think Altium saw as very attractive - seeing Upverter as a huge enabler to serve that community. Would you agree with those?
Yeah - so when we were in the kind of acquisition process with Altium, I spent a lot of time on the phone with Arum, our CEO, talking about kind of his vision for the future of Altium, and his vision for the future of electronics, and one of the things we talked a lot about was making
Altium synonymous with PCB design. And part of that is, you can't just serve the tradespeople in the mainstream - there's a million people in their basements that have ideas, that want to invent stuff, and those things include electronics. You need to be there if you're gonna be synonymous with PCB design and so - so a big part of it was that. But then also a big part of it is like, the world's changing knowledge work's moving to the cloud, is becoming collaborative you know. The Windows operating system may or may not be the operating system of the future. You look
at mobile, you look at tablets, you look at what's happening with Apple, and you know all of that. They're you know, there's a version of the world where Altium is constrained to only serving a chunk of a market because of the way we built our technology, and so I don't know, Arum obviously hasn't said any of this, but there might be a little bit of this that it is an edge on
Well I can tell you that personally, it's a really exciting place to be, because I love that. I love
that you know, I interact a lot with University students and that. But we've also gone to you know, I went to the New York City Maker Faire and to see what people are developing is so
exciting. So to be able to serve that community, and see what they come up with, is just a blast. Like I love to see it, and especially like you had a start early with IoT like it's gonna explode what we can make. And like I said, they're gonna be making it in their garage, in their basement, or their shed...
-It has to - it has to explode. Like we're talking about you know billions of devices all over the world...
-billions like capital B.
-yeah like a hundred or two hundred thousand professional electrical engineers aren't gonna invent billions of devices; we need to include a bigger chunk of the world, in the design
of these things and you know, and then
- that's not to say the tradespeople won't have their place, like of course they will - but we need grassroots, we need people building stuff in their basement. And we need it at a scale that we've never needed it at, more than now.
I know, I'm really excited about what you're doing - I'm really excited what you guys developed, and I'm so excited that that you're part of our team. So sort of wrapping up here, I want to sort of segue into what I call 'designers after hours.'
So, I don't think you have any after hours, you might get to have a beer after work once in a while. This guy's from Toronto and he's here an awful lot and I don't think you have any after hours right now, but if you did have after hours Zak, what would you do, or what do you like to do?
Yeah well my wife and I bought a house in the Canadian wilderness about 18 months ago, and so my after hours, for the last 18 months, has been turning this kind of run-down cabin into a home for my family and so I've been watching a lot of YouTube videos, and buying a lot of power tools, and trying to figure out how to do all that stuff. But when I'm - when I'm here in
San Diego, and when I'm stuck here for the weekend, [whispers] I buy a couple videogames [Judy laughs] that's - that's kind of my thing.
Okay, and here's another super fun thing about Zak. I want you to share about - his wife is also an entrepreneur. So tell us about your wife's business cuz that's really fun.
Yeah so my wife is a very successful entrepreneur, she started a company when she was quite young, that did eCommerce, and she sold that. And then she ran another company which was for angel investors, and she sold that. And then she did something with hair extensions, and
her new thing is called Sheerly Genius, and Sheerly Genius makes indestructible pantyhose.
And you can hang a human being from stuff with nothing more than a pair of pantyhose.
There is a video, we are going to share the link of Zak, hanging his wife... [Zak laughs]
from a second-floor balcony with pantyhose - it is a real thing.
It is a is real thing.
Okay, so what are the materials?
Yeah so it's heavy molecular weight polyethylene, which is what the fiber is made out of, and it's special in that it's incredibly strong, but also a low denier - or denir - I'm not sure how you're supposed to say that word - but so it's a 30 denier fiber which is what you can make hosiery products out of, but it's incredibly strong. It's like the the strongest dental floss you could
possibly imagine, and she found a way to weave this into pantyhose but also to like wrap lycra in it so that it's stretchy - but it's still sheer, but it's like incredibly and ridiculously strong, so it will never run.
-It's like 'superhero pantyhose' you guys really you're gonna have fun watching this video that we'll connect below, and we will also connect to Upverter, and EE Concierge and Zak's LinkedIn profile, if you'd like to connect with him, and any other things that we think that..
-Sounds good [laughter]
-you might want to connect to. So Zak, thanks again.
-I'm so looking forward to working with you and seeing what we kind of collaborate with and
sort of reach - reach to the grassroots-end of the design community.
-Me too .
I'm really excited about it so thank you so much for joining us.
Of course thanks for having me.
This has been Zak Homuth, is that close enough?
Zak Homuth and Judy Warner with Upverter and Altium, and EE Concierge. And we will
look forward to seeing you next time.
Until then, always stay on track.