Judy Warner (00:01.89) Hi everyone, it's Judy Warner. Welcome back to this week's Ecosystem Podcast. Boy, do I have a gem for you today, which is Tomide Adesanmi of Circuit Mind. He is the co-founder of this new startup out of the UK, and he's gonna talk about his journey from being a systems engineer at BAE Systems to becoming an entrepreneur and creating a software platform that will help take you from architecture to schematic in about 60 seconds. He's used all that experience as a systems engineer, plus his partner's deep experience in algorithm design to create an amazing tool. I think you're really gonna enjoy this one. And if you do, please remember to like, subscribe, and become part of our community. I'd really appreciate that. And it helps me bring you more incredible guests like Tumiti. Now let's jump into our conversation. Judy Warner (00:00.918) Hi Tomita, so good to talk to you across the pond. Appreciate you joining us. Why don't you take a moment and share your professional background and tell us a little bit about CircuitMind. Tomide (00:13.986) Thanks a lot for having me, Judy. So I'm Tamide, I'm the co-founder and CEO of CircuitMind. Before CircuitMind, I was an electronic systems engineer. So I used to build helmet mounted display systems and heads-up display systems for jet fighter pilots at a company called BAE Systems. It's in the aerospace and defense industry. So I left BAE Systems about five years ago right now, cause I got tired of reading data sheets. But more interestingly, you know, on one of my projects, It took us four years and tens of millions of dollars to launch a new product. And electronics development was the critical path. So on that project, I looked back and thought that we spent so much time on routine tedious tasks that didn't really need my experience. And also I just didn't do the best job I could do on the more innovative and creative and experience-led parts of the design. So... The reason why I left BAE was to find a solution to those problems. And that's how I kind of started CircuitMind. Judy Warner (01:22.706) Okay, so because you have been a systems engineer on some pretty complex electronics and that design cycle sounds awful, but it's not uncommon. So what are those specific problems besides reading, getting sick of reading data sheets? What were the other problems that you didn't like that you saw that other circuit designers had? What more specifically besides data sheets? Tomide (01:51.39) Yeah, yeah, so I can kind of try to elaborate on this. So there's this cliche. Everyone has heard the phrase hardware is hard before. And that's kind of because when the world or business people or decision makers or consumers are looking in at what does it need, what do we need to get from an idea to a product, a hardware product. Judy Warner (02:02.967) Yeah. Tomide (02:21.518) they consistently get met with the answer, this is a slow, risky journey to go from that idea to a product-ready kind of product. So the main top level problems here are that the process is slow, the outcomes aren't optimized, so engineers don't have the tools or time to kind of optimize a design to get the best potential result, and it's very risky. I've never seen a design completed right first time. So all of these are kind of top level generic problems. Every single one of us know them. But the way I parsed it out for myself is I focused on two main root causes of this problem. So the first one is that there are, I kind of divided the way I used to do design kind of categories. There's the creative design, there's experience-based design, and there's routine design. So for creative design, this is invention. This is why everybody got into electronics. This is why we wanted to be Iron Man, right? And Tony Stark telling a computer to help them to do or trying to build something new in the world. And so these are the sort of parts of your design that require deep thinking about physics. So creative design. There's experience-based design. This is... the kind of old fashioned how to take a design and make it robust and reliable for a specific application. Adding the right voltage protection circuit for an automotive application or tweaking the ESD surge protection on the design, that sort of thing. And then there's routine design. That's everything else. Choosing a microcontroller and ensuring you have the right interfaces or it connects to the right sensors, add the coupling, all of that sort of stuff. And so the first root cause of the issues in the design process is that we spend too much time on routine design, and we don't spend enough time on experience-based design and creative design. So one of the aims behind what we're doing at Circuit Mind is to automate a lot of those routine design tasks and processes and help people focus on creative design and experience-based design. So that was one of the problems. Tomide (04:39.646) And then the second problem is related to data sheets, but it's a lot more than that. I was being a bit jocular when I talked about data sheets, but yeah, the electronic design process, to do a design, what you're really trying to do, if you think about it, is you're trying to meet and balance some key design concerns. You're trying to meet the functionality of your design, the power requirements, the performance requirements, size requirements, cost availability. obsolescence, compatibility of components together. So you're trying to figure out all of these things and the way you're doing it is that you're wading through billions of data points scattered across numerous unstructured data sources, like data sheets, manufacturer websites, distributor websites, existing in-house libraries, previous designs you've done in the past, procurement databases within your company, and sometimes Excel spreadsheets. And finding the right data points, then making an optimized component choice, you know, optimized component choices and schematic design while not overlooking anything or making mistakes is very time consuming. And so this is what engineers have to fight with every single day. Judy Warner (06:02.402) So one thing I'll share with our audience. I did ahead of this, Tomedia was so kind and he showed me a quick demo. But I'll tell you the thing that really made my jaw drop was, and I'll have him talk about this later, is there's a drag and drop really cool user interface, but then there's a slide bar. an intelligent slide bar for design trade-offs. And I was like, oh my gosh. And it's actually intelligently designed with some AI, but we'll get into that later. But what I hear you talking about is trying to weigh out all these super complex trade-offs in a very analog way. And as we all know, data sheets are seldom right. And trying to... Tomide (06:55.127) That's another problem. Judy Warner (06:56.906) Well, and engineers talk about that all the time on the podcast. And so, you know, people talk about, well, let's not stand around and admire the problem, right? Let's figure out a solution. And the solution, I think is complex, but we're gonna dig into that a little bit more. So because of that complexity, tell us a little bit about your journey, your origin story of CircuitMight. Like, how did you meet your co-founder? Tomide (07:08.374) Mm-hmm. Judy Warner (07:24.338) And you can tell me a little bit, remind me about his background and how your two specialties together helped you think through how to solve these complex problems. Tomide (07:36.43) Great question. It was 2018, I just quit my job because I was on a hunt to solve some of these problems. I had the skills in the sort of electronics design and systems design area, but I didn't have the other half of the skill set, which was algorithm design, like deep, deep algorithm design. This is a very tough problem. And so I met my co-founder. on an accelerator called Entrepreneur First that brings together very smart people who are ambitious and want to start a company. And my co-founder had just finished his PhD in ETH Zurich which is one of the best labs in the world for kind of electrical engineering. So he has two masters, he actually has two masters and a PhD in algorithm design. And one of the things that I had to pull him away from was Judy Warner (08:23.886) Mm-hmm. Tomide (08:33.122) He was evaluating this postdocs at Harvard and Berkeley. And I said to him that we have, you know, a tiny window to make a big difference. And we have complimentary skills to do this. So what we did in the end was we worked together for a while, explored this problem together. I was describing how to select components, how to read data sheets, what's important. what's not important. And he was just sketching out mathematical equations saying we could try to use this kind of algorithm and we could do this for that. And that's when we started working together and we talked to a few potential customers. They came on board as early adopters and we started working with them to kind of flesh out our solution. So it was his background in algorithm design. my background in electronics design, that combination of backgrounds led to this sort of building algorithms for electronics design engineers. Judy Warner (09:42.026) I'm gonna pause. I'm gonna pause for one sec, because I keep hearing your phone. Can you turn off your notifications or something? I keep hearing. Tomide (09:51.153) Oh, okay. Let's see. Judy Warner (09:56.654) Which is no big deal, I just don't want it. I want the best product for you, so. Tomide (10:01.482) Yeah, so I know what I'm gonna do. I'm gonna put my phone on do not disturb. Judy Warner (10:07.138) There we go. Tomide (10:17.454) All right, let me know if you hear that again. I think with Do Not Disturb on, we should be good. Judy Warner (10:19.962) Okay, I will. I think I think that's right. Tomide (10:28.514) How's it going so far? Judy Warner (10:28.526) Sorry, I think it's going great and you're telling a great story and you're keeping the right balance between storytelling and enough detail on solutions. I think it's going fantastic. Okay, so now I'm just going to jump back in. I'm going to take out this piece. Okay, so I'm going to just follow up. I'm going to respond to what you just said. Tomide (10:31.84) Okay. Tomide (10:45.514) Okay. Judy Warner (10:58.354) Well, to me, that really matches up. I try to just, at least from a high level, keep my fingers at the pulse of, of trends and movements in the industry. And everybody keeps talking about design workflow and how EDA tools are going to be part of that and how AI might be part of that, as long as we get good data sources and we don't put junk in, you know, the, the engines. And so. This. makes perfect sense to me, but I also know it's hard to do. So for listeners, why don't we just go right to some of the unique features and capabilities and how you've applied those algorithms to solve some of those complex problems for listeners. Chomiday is going to share. I'm going to do that again for our listeners. Chomiday is going to put up a slide here shortly. So if you're listening. He'll walk you through it. If you have a chance, go over to YouTube and you can watch the video version. So, okay, take it away. Tell me they can't wait to hear about more details of CircuitMind. Tomide (12:09.678) Thanks, Judy. I tried to share. Judy Warner (12:13.611) whenever you're ready. Tomide (12:15.294) Oh, should I share right now or should I send you the... Okay, I'll send you the slides today. Yeah. Thanks, yeah. Judy Warner (12:17.918) You can. Oh, you can send me the slide. Send me the slide. So just talk it through. Yeah. Sorry we didn't, but you could do it. Go ahead. Tomide (12:30.51) Thanks a lot, Judy. So at CircuitMind, what we've tried to do is we've tried to focus on one part of the design process, and that is the front end of your design process. That's from the point where you get an idea to when you're done with your schematic and it's fully verified. But this does not include layout. And the reason why we were focusing there is because we believe that that's where you can get a lot of great optimization outcomes, where you can create a better product. if you do that part of your process better. So the focus of the platform is to help you speed up your process from architecture to schematic, including component selection. So right now we're looking at numbers like 1000X, the speed for some of this more routine parts of, routine design parts of your process. But that's one part. The second part is developing better circuits by optimizing your design for function, power, cost, performance, size, and all of these different things. So that optimization is key. The third part is reducing respins from schematic errors by performing advanced automated verification checks. And then the fourth part is reducing respins from availability issues by automating design for availability. So the product itself takes you from architecture to schematic in 60 seconds. So how does it work? So the CircuitMind platform Judy Warner (13:56.098) Yeah. Tomide (13:58.646) The CircuitMind platform sits alongside your current CAD tools, eCAD tools, so you don't need to kind of change your process. It just slots in perfectly. And you specify your requirements as a block diagram at a high level. So currently you're probably doing your block diagrams in Physio or PowerPoint or Word documents. All you need to do is, instead of starting in PowerPoint or even on a piece of paper, start within the platform and use a drag and drop interface to create your block diagram. It's very straight and easy. Then you can specify low level requirements on each of these blocks. Those lower level requirements are intended to capture your design intent. So for instance, you can say on a memory block on your block diagram that I want this memory block, I want the memory size to be greater than or equal to 30 megabits. You can specify that. Or I want my accelerometer to be XYZ axis accelerometer. or I want to specify the sensing range. Or you can even go as far as saying, I want my microcontroller to be from STMicroelectronics. Or you can even specify, I want the specific part number just to the rest of the schematic form. And then so you can specify that lower level detail that captures your design infant. Then you can specify trade-offs like cost, size, power, availability. And this is, I think what you were talking about, Judy, earlier. Judy Warner (15:22.552) Mm-hmm. Tomide (15:24.802) where you have these sliders where you can say, I want 100% optimized for cost or 20% optimized for power. Click and then after specifying all those trade-offs, you can click a button and the platform searches through at this point, quintillions of potential options. So quintillions is 10 to the power of 20, right? And it selects components, checks compatibility, connects them together. and generates a schematic for you, including decoupling capacitors, level translators, pull-up resistors, pull-downs, crystals, regulators. And it does all of that in roughly between 30 seconds to 10 minutes. So, you know, 30 seconds to 600 seconds. And the key to the platform, you know, that one thing that we've learned a lot from our current customers is that it unlocks a new capability called solution exploration. the idea that you shouldn't just be generating one design option. You should generate five to 10 different design options, optimize for different parameters like architectures, different architectures, cost, power, size, microcontroller tool chains, performance, stock, availability, and then explore them against each other, iterate on them, and finally choose the design option you wanna go forward with. So this is the power of using automation. you with you as the architect and the controller and the decision maker to iterate and find the perfect design for your needs. Judy Warner (17:02.582) What I love about that and you tried to articulate it and maybe you can do that better now, was you are still the driver as the creative, inventive engineering mind. So I had said to you when we first talked, it's like human in the loop, but it's not. It's like AI in the loop serving the engineering mind. And I'm like. Tomide (17:04.47) if you try to... Tomide (17:27.374) human. Exactly. Judy Warner (17:30.798) That is fantastic. So I thought of it like a, a tool chest or a kid's sandbox, like, Oh, I'm gonna try this, this and this. And it is really fast. And then you can compare. And really, I was so impressed. So yeah. Tomide (17:50.718) Absolutely. I think you articulated perfectly there, Judy. There's this concept of AI in the loop and it's a very powerful concept because it's AI is learning from humans to do something later on and improve their processes. In this particular case, it is these tools, these algorithms, including some parts that are done with AI, not all of it. these algorithms actually serving within your loop, your loop being your design process. So what this platform enables you to become is now an architect, a verifier, a solution explorer. And of course, there's still some parts that it's not gonna be able to do for you. It's not gonna be doing the most creative parts of your design process. This is focused on routine design. Judy Warner (18:24.01) Yes. Yeah. Judy Warner (18:33.144) Mm-hmm. Judy Warner (18:46.155) Right. Tomide (18:46.334) If you're doing a new waveform generator, it's not gonna be able to do that for you quite yet. If you're doing things that are more complex, that's where your creativity is needed. If you need to get to be to automotive qualification, it will, and you need to add some more parts of your circuit to get it to be reliable, you need to do that yourself. But this platform can help you get answers and get to a schematic on, let's say majority of your design process very, very quickly. And that's kind of one of the things that I, like I use the platform every single day and I'm grateful for. I don't really need to go and open data sheets anymore, but I can get very, very quick answers. Judy Warner (19:15.841) Right. Judy Warner (19:26.494) Right. And that is part of engineering, of course, to do some of that. But it sounds like you've taken a lot of that pain out of it, which I think is very exciting. So you mentioned that early on you sort of had, you know, beta users and, and partners that were giving you real world feedback, but these just weren't You know, software engineers, these were people making products. So tell us a little bit about that and how you really proved out. I mean, you're using it. You're an engineer, but how did you run sort of your beta and how you developed this and then what was the kind of feedback you got? Tomide (20:11.438) That's a great question. So we knew from day one that we're building software for professional engineers who develop products that go out to thousands of people. So like you said, from day one, we needed to develop this to meet the standards that they expect. And so what we started to do in our early days was we started building, we started by being a design consultancy. I... and one of the early members of CircuitMind called Gary, who was, you know, he spent 20 years in the industry. He was also at BAE Systems building helmet mounted displays, heads of displays. He was a leader on many of these projects. We take on design jobs and we put it into the platform and say, into the early platform and see what the output was and say, can we actually use this in our project? in this project. And until it got to the point where the answer was yes, we didn't let anybody else touch the platform. And then once it got to the point where we, where, you know, the answer to that question was yes, we then started onboarding real engineers, sorry, external engineers onto the product to test and give us feedback. Judy Warner (21:11.36) Mm. Judy Warner (21:19.298) Hmm. Tomide (21:39.318) and we wanted them to use this on real world projects. And so we had some of our early kind of success stories from some of those beta testers that iterated with us. I think the key in that kind of beta testing part of our journey was making it usable because when we were using the platform as a consultancy, we didn't really care about the user experience, but when we put it out to other people, they were like, oh my God, Judy Warner (22:05.822) Mm. Tomide (22:09.374) This is painful. But we worked through that. And right now, I think the user experience is one of the things that I'm proudest of in the platform. And so we moved on from beta testers to trials with companies, paying customers over the years. And we're just iterating step by step. And then we've gotten past the part of our journey where someone that's used the platform on their own has taken a product to prototyping. And and early production. So now we're kind of beyond the point of, you know, just validation of the outputs of the platform. And it's now more about the bigger value that it creates for our customers. Judy Warner (22:51.566) So I heard you say then you have paying customers, like you've gone through all your proof of concept, you fixed your UI and your UX. And so what has been the feedback? And is there one maybe really memorable story where you felt like, aha, we got it. Tomide (23:12.17) Yeah, so the feedback, I think in general, so right now we're, like you say, we're serving customers in industrial electronics, design services and EMS, consumer electronics. We're serving companies in small startups in the sort of carbon monitoring space and more. And we're kind of growing across many different kind of industries. And... I think the biggest, the feedback that stood out the most for me that I'm really proud of is this new capability of solution exploration. So for instance, one of our customers, product development firm, they developed a smart air quality monitor for one of their customers. So this is a company that does designs for others. They went from idea through trade-off or solution exploration. to component selection and schematic design in our platform in about half an hour. They explored about five different designs and the best designs that they came up with were compared against the design that they had done manually and they saw 15% reduction in bomb costs, 22% longer battery life, 43% decrease in size, all parts available. and the prototype that went out to the customer was done without any respins. I think that's unbelievable. Judy Warner (24:44.163) That sounds like a unicorn in the... You know, that doesn't even sound real to have those kind of improvements. Tomide (24:49.416) Ebb. 100% and I don't even want to quote the time, you know, I can talk about X thousand, you know, in speed, speed improvements. But I don't even want to talk about that because for me, the better design, you know, the ability to do that, that design and kind of balance these things a bit better, like the battery life and cost and size of the design. I think that was, that felt more powerful to me. But of course the speed improvement is still good. You can accelerate time to market, but we know that there's more and more in the process. Well, that we spin as well. Yeah. Judy Warner (25:31.858) Well, I wanna press you on that. I'm gonna press you on that because that was the first thing I wanted to ask you about, right? Because early on you said, you know, BA might take four years. So, I mean, we're not gonna hold your feet to the fire. And I know every, and everybody that's listening or watching, they know every electronics product is completely unique. But like, Tomide (25:45.928) Mmm. Hehehe Judy Warner (26:01.23) For this customer, like how much faster? Just that they didn't have to re-spin is miraculous. Much less parts availability and all the things that you mentioned. So come on, tell us, Tommi, we want you. Tomide (26:09.228) Yeah. Tomide (26:14.25) Yeah, so we don't know the exact number because they just went forward with the design we were going and we didn't kind of have a benchmark to go again. They weren't going, they weren't taking for their manual process and also doing our process. They did an initial bit that we could benchmark against, but they didn't do. But I would estimate it's around 20% because Judy Warner (26:24.545) Okay. Judy Warner (26:30.612) I see. Tomide (26:40.238) Obviously, there's other parts of your kind of hardware development process. You have mechanical, you have firmware development. So you still need to make sure that those things kind of are completed and go on. They still take the note. But the point is that anything where the electronics process is the obstacle or was the critical path or, you know, created... Judy Warner (26:54.914) Right. Tomide (27:08.287) or sunk resources that was eliminated and you could kind of shorten that time to market by let's say 20%. That's what I would estimate. But the answer is I don't know 100%. Judy Warner (27:10.818) Mm-hmm. Judy Warner (27:19.192) Wow. Judy Warner (27:24.122) Yeah, no, I get it. But I think, even if you said the price was lower and you did one spin instead of three, you know, and prototyping with physical hardware instead of sort of getting this virtual sandbox version of your circuit design schematic in that kind of speed is. So are there other tools out there that Tomide (27:34.262) Yeah. Incredible. Judy Warner (27:53.522) I mean, you mentioned to me before that this obviously doesn't eliminate your EDA tools. It helps you work better and more intelligent with your EDA tools. But are there other competitors out there or are you sort of in a blue ocean mode and trying to create something completely new? Tomide (28:12.826) Um, so we are, yeah. So, so first of all, exactly. It works alongside your current E-CAD design process. Your E-CAD design process should just be focused on that creative bit and the places where you need to add your experience. And also layout, of course, it doesn't do layout. But in terms of competition, there's not much competition out there. There are a couple of companies that are, um, Judy Warner (28:30.603) Right. Tomide (28:40.802) kind of working on what I would call new age development tool ideas that some of them are helping with trying to automate layout and routing, which is an age-old problem. I know engineers hate auto-routers, but there are some companies that are kind of trying to do that. But there's, I would say, one or two companies out there that are trying. Judy Warner (28:57.987) Mm-hmm. Yeah. Tomide (29:09.546) in the front end of the design space. But everybody's just doing it very differently and attacking it in such a different way that we're not even speaking to the same customers on the same path, I would say. Yeah. Judy Warner (29:15.469) Right. Right. Judy Warner (29:22.218) Yeah, okay. I see. Okay. Well, it seems to me that one, your experience as a sort of high end complex using complex systems, engineering has done a lot to inform your choices. Um, where are you in regards to like a public launch? Like I accidentally found you online cause you presented somewhere and I tried to backtrack and find out where that was, but I feel like I found a gem and It's like, I haven't seen you guys sort of across the industry or news. I know you've done some trade shows. Like where are you in that process? Tomide (30:06.254) That's a good question and I'm sure my investors would echo that question. Hey, where are you? So we're out there. You can go to our website, you can request access, we'll do a demo with you, we'll introduce you to the platform, you'll get you on the platform. We have professional engineers and engineering teams on the platform today and also directors and managers that reach out to us. But you're right in saying that we've not. We've not gone out, put adverts out there and done like a big launch. So we plan to do a proper public launch probably within the next three months. But we've been hyper-focused on just being in the hands of engineers during real products. And now that we've done that, Judy Warner (30:53.268) Okay. Judy Warner (31:03.862) Hmm. Tomide (31:05.346) repeatedly. We know what the process is. We know what they like, we know they don't like that we can, we can kind of scream and shout out more about you know, what we're building. And then we plan that in the next three months or so. Judy Warner (31:16.331) Right. Judy Warner (31:21.198) Yeah, well, I like it just because I know a lot of companies, particularly software companies, can get out ahead of their skis where they're not done developing or, you know, it'll be, it'll be done soon, but we're gonna do a big splash. And I know my engineering buddies hate that, you know, because it feels disingenuous. And you guys, you engineers, like, give me data, or I'm not gonna like... Tomide (31:41.603) Hehehe Judy Warner (31:50.698) They don't believe marketing people. So I love that you kind of did that. Although your investors may be a little frustrated. I think it'll pay off probably. And so I think you're being measured is probably gonna serve you, but we shall see. Well, this has been a great conversation. I want to encourage our listeners to go look at their website. There's some, they have a pretty cool website actually, and you can see some of it and go check it out. Tomide (31:59.085) Yeah. Judy Warner (32:20.382) You, your mind will be blown, but more than that, I go do a demo. I mean, I'm not an engineer. I know enough to make me dangerous, but I was absolutely mind blown on the capabilities and it really did turn. I think when we did the demo to maybe turned out, I don't know, four or five. Uh, sort of digital twins of the same design while we were just talking. So encourage you guys to do that. Um, to me, where else can people. You mentioned to me earlier you've been at some trade shows and that you're getting ready for a public launch. Where can people sort of keep track of Circuit Mind and stay in touch with you? Tomide (33:00.17) Yeah, so main place, go to our website, www.circuitmind.io. On the website, you can either request access to the platform where we'll, you know, if you request access, we'll take you through a demo and then we can go through trials and get you on the platform. Or you can actually sign up to a newsletter if you just wanna keep track with us, you can sign up to a newsletter. Or you can go to our LinkedIn page. Judy Warner (33:25.18) Okay. Tomide (33:29.258) Just search for Circuit Mind on LinkedIn. We do regular posts on where we're gonna be, what trade shows we're gonna be at, webinars. We do webinars as well. So if you sign up for the newsletter, we'll send you some information on webinars, but you'll also see webinars on our LinkedIn page. Judy Warner (33:35.943) Okay. Judy Warner (33:45.734) Okay, I'll grab those links from you and I'll put them in our show notes. Well, Tamini, thank you so much for your time. I know it's evening there in the UK and really I am authentically super excited about the work you've done and just congratulations and hopefully we'll be able to meet in person and I'll be able to get to see Circuit Mind, you know, at a trade show. But thanks again so much for joining us and sharing with our audience the amazing work you're doing. Tomide (34:12.686) Thank you very much as well, Judy, for having me. And thanks for kind of having these conversations with me, starting from a LinkedIn message. Yeah, looking forward to having more conversations with you and your incredible community. Judy Warner (34:21.1) Yeah. Judy Warner (34:26.882) Definitely. Well to our listeners, I trust that you enjoyed this conversation as much as I did. Make sure you go check out the show notes. I'll get all those links from Tomide Day. Until then, remember to always stay connected to the ecosystem. Ba-boom! That was a good podcast. Tomide (34:44.525) Thanks for watching.