CircuitHub aims to accelerate electronics manufacturing to software speed

0:00Mike Butcher:Welcome to Path Founders with me, Mike Butcher. We like to look at the code, the capital and the consequences in the startup ecosystem. Today, I'm joined by Andrew Seddon, founder and CEO of CircuitHub. CircuitHub is tackling one of the less glamorous but increasingly critical bottlenecks in tech, actually making hardware. While AI and software tools are making it easier than ever to design new devices, electronics manufacturing is still often offshore, manual and built for mass production rather than rapid iteration. CircuitHub's answer is an automated software-driven manufacturing system where engineers can upload design files, order circuit boards online and then have them produced in days.

0:48With research and development routes in Cambridge in the UK, production in Massachusetts, over 2 million boards delivered and plans to expand factories across the US and Europe.

1:01Mike Butcher:The company sits right at the intersection of hardware, AI, supply chain resilience and Western re-industrialization. Andrew, thanks for joining PathFounders. Pleasure to be here, Mike. so uh do you want to give us a simple version what is circuit hub and what's the problem you're solving for hardware teams yeah so we make it really easy for electronics engineers to get circuit boards made so if you're an electronics engineer you can upload the design file for your circuit board to our website we give you an instant quote from one to ten thousand units so we give you a lead time we give you a price you can then click a button pay via credit card and we then automate that back-end manufacturing process to get those boards out to you in record time.

1:44So the big thing that we really sell is speed. So if you are an electronics engineer at say like an electric car company, VR headset company, drone delivery, etc., you're designing circuit boards to go into your device and you're looking to have rapid design iteration cycles. So traditionally in the industry, it takes about two to three weeks to get an assembled circuit board produced. And we're able to do that in two to three days, thanks to the kind of software and automation that we've applied to the problem.

2:15Mike Butcher:And, you know, is this sort of thing done in China or in other countries, or are you just, have you invented something brand new? Yeah, I mean, it's an old industry. So people have been producing assembled circuit boards since, I think it was World War One, actually, they kind of came out the radar system design, actually, in the UK. So the industry has been around for a while, and in various different guises so in the uk for instance we used to have a pretty strong electronics industry all through the kind of 80s 90s and then going into the year 2000 it really kind of dissipated or went over to asia and we've seen the same thing happen over in the us as well so in the year 2000 there were about 2000 pcb fabs in the us and that is now down to under 100 so there's been this sort of dramatic drop in the number of places that you can go to to get circuit boards manufactured in the west now so yeah the vast majority of electronics manufacturing is done over in China as of as of right now or over in Asia at least but we're seeing a sort of a real renaissance in demand for domestic manufacturing and that's kind of driven by a few different factors a lot of it is the these kind of innovative like tech startups, hardware startups that are coming online, they're kind of looking around at their options for their manufacturing and they're thinking, okay, do I really want to put this massive divide between my manufacturing process and my design process, right?

3:52Because if you do that, it slows down your iteration cycles. So if you have a sort of a factory on the other side of the world, you're trying to like rapidly innovate on your product design that's going to slow you down and so a lot of them are kind of looking around at domestic options and they're seeing the fact that the the industry has just been decimated over the last 20-25 years they're not seeing any great options there and they're then thinking okay well we're going to have to vertically integrate instead and so you're seeing some of the larger hardware companies uh there's a really good example of this with spacex actually in Bastrop in Texas.

4:27They've started producing the Starlink base stations at pretty high quantities, kind of consumer electronics quantities, in a fully vertically integrated, wholly owned electronics production facility. And they've obviously, they've highly automated it. They have a bunch of, you know, really, really well-trained people doing that as well. But it's the sort of project and manufacturing line that you typically, say 10, 15 years ago, would have seen subcontracted out to to asia and they've taken the decision to do that uh domestically in the us and i think the primary reason for that is so that they can kind of maintain control of that manufacturing process they can feed back the learning from the manufacturing process into the design side and kind of go through this like rapid rapid iteration loop and so what we're um you know that's great if you if you have uh if you're a spacex and you and you have the the sort of known demand for your product but a lot of startups are not in that position they're kind of iterating on the product they're iterating on the demand signals simultaneously and so you don't necessarily want to be expending all the capex and time to spin up your manufacturing facility and so where circuit hub comes in is we offer this uh this kind of shared next generation manufacturing facility that is way more advanced than anything else that you would sort of typically see in the west uh but you don't have to build it yourself you know we we've spent uh you know all of the years on the research and development we've invested all the capital uh equipment and then you just kind of get access to that uh as a as a hardware company well okay it sounds like you're

6:04Mike Butcher:spinning a lot of plates there um and part of the reason we're talking to you now is that you've raised 28 million dollars uh with money from plural and others um what uh interests them in about you yeah so i mean plural is a kind of fascinating example i think of uh a sort of forward-thinking vc and that they've they've gone into the hardware deep tech side of things actually relatively early they did it a few years ago kind of before the industry has shifted into wall to that and so they have a pretty wide portfolio now of existing hardware companies that are all kind of consumers of manufacturing services right and so a very good way to invest as a VC is you look around at what are your portfolio companies using as products and you sort of you're looking to build an ecosystem there right like so so your your sort of next generation hardware company you if they're consuming a manufacturing service okay there's probably something interesting about that.

7:06I think that's what they saw in CircuitHub is the fact that we're serving these next generation hardware companies. So yeah, it's great to be kind of part of their ecosystem.

7:20Mike Butcher:So are you saying that you're going to be serving some of their portfolio companies? Yeah, I think that's the idea. I mean, they have a portfolio in kind of space, they have fusion companies, they're starting to do some defense investments now as well. So just to kind of give you a background to the structure of the company, I originally started the company in the UK, but we did go through sort of Y Combinator. We raised money over in the US initially. We've been historically split about 50-50 between the UK and the US, sort of primarily Massachusetts and the vast majority of our customer base is actually based over in the in the US right now.

8:03And so I think part of the the plural investment and certainly where you know some of some of the capital is going to be going is to be expanding our facility initially in the UK but ultimately Europe as well so that we can start to serve these kind of UK European customers in the same way that we're currently doing for the US customers.

8:22Mike Butcher:And what stage is that 28 million? Is it a series A or what? Yeah, it's a series A. It takes us to give or take about$45 million raised overall. So we've done a few kind of seed fundings prior to the series A. As you can imagine, getting factories off the ground is pretty capital intensive. And so the kind of stages, I think ratchet down a little bit for hardware companies. This seems to be part of a much bigger kind of geopolitical shift in reshoring and nearshoring. And Plural is talking about this in terms of an infrastructure shift. Would you agree? I think so. I mean, you know, there's two parts to it, I'd say, broadly.

9:09There's the kind of perception on the geopolitical side of things, right? Like if you are a company, you're just looking around commercially and you're thinking about the security of your supply chains. Are you going to be taking bets on stringing those supply chains out across the world, frankly? And there's been various different shocks to the system over the last few years that have kind of made that a bad idea. So I think we saw this really start and kick off during COVID. right so if i think back to 2020 obviously there was a sort of initial shutdown of the economy but when that came back in the second second half of 2020 we saw a massive uptick in uh in demand and that was really off the back of hold on you know our supply chains have just been shut down for four or five months we need to figure out some sort of resiliency here um some untangling of supply chains some ability to kind of get it more domestic and get some additional control over it and then since then uh we've obviously seen uh various different uh shocks geopolitically um i think last year uh towards the sort of the back end of last year on on the um import duty side of things over in the us was was pretty significant uh so consumer electronics being imported from china was hit with i think at one point in time it was 80 percent uh import duties which uh obviously completely breaks the economics of having a supply chain based over in Asia.

10:36That has dissipated down a little bit now and they're kind of back to a more sensible level, but we're not in the same environment that we were a few years ago. So I think that is definitely contributing. But then the second half of this, if you're just an engineer, kind of boots on the ground and you're trying to rapidly iterate on your design, you're not necessarily thinking about the whole sort of geopolitical climate that's going on. You're just thinking about how can I get my product to market as quickly as I possibly can. And for kind of a hundred little reasons, the decisions that you make there now, I think, are quite different to the decisions that you would make 20 years ago, right?

11:20Like you're looking at all of the different points of friction of do I offshore? Do I build onshore? and I think the tables have just turned there to make it much more both economically viable to produce onshore so thanks to increasing amounts of automation you know we do a ton with with robotics we take a lot of human labor out of the process that just that makes it more economically viable to produce onshore there's the you know communication time zone aspect of things is is just incredibly important and then increasingly ip protection so you know the uh i think there was this misunderstanding that you could uh you could sort of design in california and you could make in china and that the ip would would accrue to the to the design side of that equation i think in practice that the ip ends up accruing to the manufacturing side of that equation right and and so companies have seen this play out over the last 20 years and you know if you if you're coming up with like a you know a fusion reactor or you've got some like innovative drone delivery product uh electric vehicles etc you know you really got to stop and and take stock of do you want to be uh you know teaching somebody else how to make your product and so all those factors combined i think are kind of leading to this this overall shift so in a way we're sort of moving from

12:40Mike Butcher:the debate around sovereign ai to the direct a new debate around uh sovereign circuit boards almost yeah i mean i kind of think that i even think of the ai boom as a hardware boom right like if you think about where is all that capital actually flowing okay you know it's flowing in from the you know from the pension funds ultimately into the vcs and into into the startup companies and if you look at where that's being exhausted out to it's all into physical infrastructure right like it's going into utilities um you know a lot of grid grid and power build out it's going into data centers which is which is chips on circuit boards so so it's a real physical infrastructure build out and i think you know a lot of the the kind of margin that once existed in the world of bits is being compressed and uh and a lot of that capital is now being kind of diverted through to the world of atoms.

13:35So yeah, it's an interesting time to be a hardware company right now. And you know, the AI side of things is really throwing lots of things up in the air for us.

13:49Mike Butcher:Obviously, the two hottest areas right now for investment are AI and the other is defense. Is that also part of your sort of strategy? Yeah, I mean, on the AI side of things, that kind of weaves through everything we do in many different ways. So the actual manufacturing process is very much enabled by AI. We have quite a bit of on-site compute now. We've got over 100 kilowatts of compute sort of sat next to our manufacturing system, which does real-time inference and kind of enables the type of manufacturing process, a flexible manufacturing process that wouldn't have been possible a few years ago.

14:31So extensive use of kind of learned models within the manufacturing process itself. We're starting to use actual LLMs and agentic systems to do overall control of the factory system as well, which has been a very exciting development of the last couple of years. So that's enabling kind of further labor reduction, increasing consistency of the manufacturing process and reduction in error rates. so using AI directly on the factory floor and then in terms of our own internal development process we've seen roughly about a 3x increase in developer productivity so the vast majority of our team is actually software engineers sort of counterintuitively and you find this in a lot of hardware companies is you have a sort of base of mechanical engineers that are doing all the physical real world stuff but then you still need a kind of ton of software engineering to sit on top of that and that's that's the same for circuit hub and so you know we found we found that ai has just been a massive accelerator for our kind of own internal product development and then um in terms of our customers i think ai is just going into all different physical real world systems right you know fusion companies are using it using it to control fusion reactors There's drone deliveries becoming actually feasible and works now as a result of AI systems.

15:54And those deployed systems are typically going to be a chip, maybe an NVIDIA chip, an AMD chip, that's mounted onto a circuit board that's fitting into some other sort of plastic or metal product. And so that's something that we can get involved in. And that general robotics boom has been sort of really beneficial for us in terms of kind of a growing customer base and the types of customers that we really want that are kind of forward thinking customers. So, you know, that's on the kind of AI side of things. It's been a real boom for us. On the defense side, we have traditionally been sort of, I would say, non-defense focused.

16:39And the reason for that is over in the U.S., actually the only remaining electronic supply chain is defense. So pretty much all non-defense work went to Asia. And then the remaining supply chain is kind of defense based. And so we, as a first step for our offering, we wanted to make sure that we offered something that didn't previously exist. You know, now that we've been able to do that, I think, you know, we are looking around at some of these kind of more modern defense companies. We do count some of them as customers already for sort of non-critical technologies, ground technologies and such right now.

17:21but I think that is something that we'll look to go look to address going forward.

17:28Mike Butcher:Well, it's clearly a big vote of confidence by Plural putting 28 million into you but thanks very much for joining Path Founders and Andrew Seddon, Founder and CEO of CircuitHub. Thanks Mike, much appreciated.