Judy Warner (00:01.597) Hi Zach, thanks so much for joining us today. I'm delighted to have you as our guest and excited to talk to you a little bit about breadboard today. Zachary Feuerstein (00:09.76) And nice to see you as well. Judy Warner (00:12.317) Well, it was very fun meeting your team last week at PCBWES and I thought our audience would love to hear about the very interesting things you're doing with AI and LLM. So before we get started, I want you to tell us a little bit about who you are, what your background is and what led you to start Breadboard. Zachary Feuerstein (00:30.722) Sure, yeah. So my name is Zach. I traditionally have always been involved in electrical engineering and computer engineering. My father is also an electrical and computer engineer. So I grew up all around it, whether it be our TV breaking or something in our house, always to maybe my mom's chagrin, right? Like we were responsible for fixing it instead of just buying a new one. Got my education in the space. But realized when I was like thinking about my first job that hardware is really hard as probably all your viewers know way better than I do. And software was very enticing at the time and decided to start actually by first company in the software space as a senior in college, started that company and really, you know, started with some successes and some failures, a bunch of companies with the most recent one. working a lot with medical device companies actually. So got to see hardware again. And then during the supply chain crisis, a few years ago, my fridge broke and I wanted to fix it based on all the lessons, you know, my father had taught me of like, fix things, don't just buy new ones. And when I went to source the motherboard for it, they told me anywhere from six months to three years to get it. And I was like, something is going wrong here. Judy Warner (01:52.839) What? Zachary Feuerstein (01:54.412) Yeah, it's crazy and we should definitely fix what's happening there and been working on the electronic supply chain ever since. Judy Warner (02:02.77) So tell us exactly what problems you're trying to solve with breadboard and how does it work? Zachary Feuerstein (02:10.86) Yeah, so Breadboard is an electronic supply chain platform that empowers engineers and manufacturers to quote, purchase and track every part they need to build the products that they're working on. Right. And what we've done is we've simplified the traditionally manual and fragmented process of, you know, electronic sourcing and manufacturing. And the way our system does that is it integrates with real time data from all the different manufacturers, distributors, and different players in the space on every component you need. So covering technical specs to compliance to availability like stock and pricing and integrates that to give you a very, very simple place where you can upload your requirements for what you're trying to build and go all the way from those requirements to actually getting those parts delivered to your facility in a simplified and automated fashion. So that's kind of what we do. really that process helps you not only with the manual side of all the hours you're spending and days getting that data and processing it, but also reduces the risks so that in the cases where parts are not available, how are you going to find a secondary source for it and making sure that there are no delays in your design to manufacturing process. Judy Warner (03:32.733) So does it give you that forecasting information like end of life on parts that you can bake that in or how does that work? Zachary Feuerstein (03:41.57) Yeah, so not only does it provide you that data, so for example, part lifecycle, like whether a part is obsolete already, or maybe that it's going obsolete and not recommended for new design, but what we also do is recommend parts that could be a good replacement. So instead of your sourcing team having to go back to the engineers and say, hey, we can't find this part, you're going to have to redesign it, right? Judy Warner (03:53.707) huh. Zachary Feuerstein (04:06.53) they can at least come with suggestions, right? And we'll provide those form fit function alternates for, you know, the different parties who are looking to procure these components so that they have an alternative solution instead of just raising their hands of the problem. Judy Warner (04:23.645) So that's kind of an, at least with some of the other companies that are trying to take a stab at the same idea, that seems like a differentiator. Is that for your solution, having that kind of proactivity? Zachary Feuerstein (04:40.268) Yeah, I think that's one big piece of it, right? Is making sure not only do we give you the data, but we give you a solution to that problem that's there. But also, right, what we've noticed in this space and really where we've doubled down on, and I think a lot of people are seeing is the complexity of all this data that's going back and forth. So whether it be the distributors or suppliers responding in data in weird formats. Judy Warner (04:49.459) Mm-hmm. Zachary Feuerstein (05:08.49) or it being that your bill of materials that you're working on has lots of alternates in different fields. Really where we specialize is taking that unstructured data and putting it into a structured format so that you can not only get those suggestions that we were just talking about, but also take action on it to make sure that you can move forward instead of spending that time cleaning up all that data by hand, which unfortunately we see lots of people still doing in this industry. Judy Warner (05:36.015) Yeah, absolutely. So, as I was at PCB West last year, and again, I've been around for a very long time and been to that show, I don't know, probably 20 times, lots of times, it was so remarkable. What stood out to me is how much people are talking about AI and LLMs and how that's going to affect the space and how many different opinions or people are coming at it from a different angle. And I know that's a technology you use. So when you talk about gathering and structuring that data, how are you using maybe LMS, AI, and some of the newer technology to put that into a structured usable way and create an intelligence solution for design engineers who are our audience here? Zachary Feuerstein (06:31.052) Yeah, so that's one technology that's being used in, I think more importantly, it's how does that entire stack work together, right? You wouldn't want to just throw your bill of materials or your Gerber file at chat GPT and say, Hey, good luck. Like, tell me what I need to do with on it. It's not trained for that. It's not ready to deal with it. really what we've built is that workflow that first takes a document from any format. That document could be a bill of materials. Maybe it's also an RFQ response, or maybe it's an email that says about a change order that's happening about a specific PO, and translate that into a structured format that an LLM or something like that can understand, that then it can recommend actions based on. So for example, if you have a drawing of a wire cable harness or something like that, how do you extract first the bomb off that? Judy Warner (07:16.976) I see. Zachary Feuerstein (07:27.406) Then how do you start asking questions about that build materials to it? And then from there, how does that translate into data that can be actioned up on into our system? So really the thought behind it is there's a lot of data that's happening in, you know, maybe not the simplest of formats. Like for example, that email has a change order. How are we using that to suggest an action that can be taken upon in our system? So. For example, you submit a purchase order, you say, hey, I'm buying this part for 81 cents. The supplier responds to you and says, hey, actually, I can't get it for 81 cents, but I can get it for 93 cents. Do you still want this part? In the traditional sense, you would have had to go into your email, find that email, realize what the supplier is saying, go into your ERP system, get that data, put that back into your ERP system, reissue a new purchase order. And then have that new version versus an art system. We read that email. We understand what that means. We generate that change for you. We issue that new PO. We track that version history so that even when you know, on the financial side, you're trying to tie out that PO to an invoice, right? It's all there for you and just ready to go. Judy Warner (08:44.275) So it's able then to, it, you're talking about which another company I talked to said again that they're able to with permissions of course, able to gather that data from emails. Are you actually connecting into your client's ERP or like how do you close that loop once you've gathered that data, structured it? Like how do you get to that point and Zachary Feuerstein (09:11.468) Yeah, so we're- we're an automation layer on top of that ERP system. So whether it be us reading data out of it for inventory or historical pricing, or us writing data back to it for our POs and order statuses and everything like that, we're not gonna be replacing that system. We're gonna be sitting on top of it and then allowing you to update that without manual intervention. And we do that both on the ERP side, but also to be honest on the supplier and the distributor side, Judy Warner (09:25.828) Okay. Zachary Feuerstein (09:43.97) where if they support it, right, we're going to be issuing those POs through APIs and through transactions that way, so that there's not the manual work on their end to process all of it, which can cause delays as well. Judy Warner (09:50.491) Mmm. Judy Warner (09:57.475) Amazing. That's pretty interesting. So that's how you can do the end-to-end solution then, right? You're covering all those steps to the point of a PO, which is an interesting solution I haven't heard before. Zachary Feuerstein (10:12.746) sure yet, whether it be the suppliers work with API or EDI, or if they just simply want it to be done over email, right, our system can handle those levels of integrations. Judy Warner (10:24.687) Okay, well let's go back to those differentiators. What are, you know, you've seen them too. There's lots of different people coming at this from various different angles trying to solve a similar problem and maybe a different way. So what makes breadboard unique? And then let's just start there and then I got a secondary follow-up question. Zachary Feuerstein (10:49.23) For sure. Yeah. So I think there are different people that purchase different ways. Our belief is that while people want to automate this process, make it easier, reduce the risk, they need to do it in a process one that they're familiar with and with the existing suppliers that they're already working with. And really that's where we want to make sure that we can support. So when you're issuing those POs, right, We're not, you're not issuing those to breadboard. You're still issuing those directly to your same suppliers. You're dealing with the financing side that way. If you have different facilities that have different requirements in terms of where things are shipping and to who, and you know, maybe even have different like rates of attrition, right? Of extra components that you need to order. Our system allows you to configure all that on the fly. So. Judy Warner (11:21.863) Mm-hmm. Zachary Feuerstein (11:39.308) while you're maintaining your existing processes and configurations, which can be quite complex, right, in electronics manufacturing, you can automate on top of that instead of replacing that with some other solution and then having to rely on that sole partner. It's more of adding that automation layer on top of your existing processes. Judy Warner (11:46.439) You bet. Judy Warner (11:59.781) And do you have, is the CM sort of in this workflow? Zachary Feuerstein (12:05.902) Yes, we both see CMs and OEMs as our customers. We see more complexity, to be honest, on the CMs. And we have over 50 % of our clients are these CMs who are dealing with 20, 30, 50, 100 clients, which all have their own configurations and settings. it's really whoever is buying the components. see both some OEMs are buying their own components and then either manufacturing it themselves or consigning it to the contract manufacturers. Judy Warner (12:16.114) Okay. Judy Warner (12:21.575) Mm-hmm. Zachary Feuerstein (12:34.274) And then a lot of contract manufacturers who are doing turnkey work who are responsible for purchasing those components and are using our platform to do so. Judy Warner (12:44.029) So Zach, how long has Breadboard been around and what kind of feedback are you getting from both CMs and OEMs? Zachary Feuerstein (12:54.892) Yeah, so in terms of breadboard, we've been around just over two years now. We launched our platform about a year and a half ago and we're getting great reception. think, you know, initially there are people who are always, you know, a little hesitant, rightfully so with bringing on new software. They want to make sure that one is it's actually going to be able to do what it says it can, but even more importantly, that it's working with their existing team and not just replacing them. But, you know, once typically we're showing people the software, explaining what it can do and how much of an improvement it can make in their jobs so that they're doing those value add activities like supplier relationships and managing those and working on, you know, improving their structures instead of doing the manual work. People absolutely love it and we're getting great reception there and we're just constantly trying to improve our offering. When we started the platform, we were purely working on quoting. Now we do purchasing and supply chain monitoring and we're thinking about lots of other ways that we can improve and make the process a lot easier from, you know, all the way from design through, you know, delivering that product to the customer, you know, at end of the day. Judy Warner (14:06.781) Pretty great. So about how many active customers you have now on your platform. Zachary Feuerstein (14:13.314) Yeah, so we have over 60 customers that are using our platform every day, ranging from smaller startups on the electronic manufacturing side to top-top contract manufacturers in the world who are doing hundreds of millions or billions of dollars worth of transactions. Judy Warner (14:30.897) Well, that says a lot because as you said, that's a really complex because they're handling so many customers that to me to give some weight to what you're doing. Okay, now I want to play devil's advocate if you let me. Zachary Feuerstein (14:47.086) Yeah, of course. Judy Warner (14:48.317) So, some of my peers that have been in industry for a long time that either do PCB layout, hardware engineering, and I know you've heard it many times is they're all like, this will never work, it's too nuanced. If you don't want to read data sheets, then you should be a design engineer. So, what's your response? that kind of thing and I get what they're saying. At the same time, I have to say honestly, I think I'm standing in the camp of you, Zach, and all these young entrepreneurs that grew up coding in elementary school and figuring out software. I think you will disrupt it, but it is very complex problem. So what do you say to people who kind of assert that kind of voice in the industry? Zachary Feuerstein (15:48.92) I think there's a lot of truth to it, right? think there's, computers are really, really good at taking a very highly defined problem and running through that if it has all those requirements there. What it's not really good at, even with the greatest technologies in the world, making judgment calls. Humans are really, really, really intuitive and really smart at making judgment calls that a computer, which at the end of the day is just taking inputs and making outputs from it. is not as good at, right, at making those inferences. It's getting better, right? Don't get me wrong, but it's not quite there yet. And really what we think is what you get perfection out of, or as close to possible, call it asymptote in towards perfection, is the blend between humans and technology, right? So yes, people will say, hey, you need to read the data sheet. But do you need to read that by hand, right? Like, you know, Judy Warner (16:24.787) Mm-mm. Judy Warner (16:38.247) Mm Zachary Feuerstein (16:45.666) Probably not the same way, right? You wouldn't say you need to grab the textbook off the shelf to read that physical data sheet. And now it's in a PDF. You wouldn't ever question that. Well, are there tools that can help you search through that data sheet better than a control app, right? To get you that result. Yeah, there are. and really, you know, at the end of the day, you want that person saying, yes, this is what I want, right? And approving or denying it. But can we get them that information to make that Judy Warner (16:56.529) Move. Judy Warner (17:01.777) Right, right. Judy Warner (17:10.151) Mm hmm. Mm hmm. Zachary Feuerstein (17:14.475) decision a lot faster so that they can spend their time making those judgment calls instead of doing the manual work to get the data to make that judgment call. So I think there's a lot of truth to it. We're never going to, and we don't want to, get humans completely out of the process. I don't think that would be a good thing. But how can we let them do things that are interesting problems to work on, take their brain power to do it instead of spending 80 % of the time on the boring and manual work? and only 10-20 % of the time that the projects they actually want to work on and flip that around. Judy Warner (17:50.675) Have you ever collected data on that, say how much time engineers spend on whatever, scrubbing bombs and reading data sheets, or maybe from your clients that gathered some data that said it's saved us this percentage of time? Zachary Feuerstein (18:09.794) We definitely do. We do it mostly on the manufacturers versus the engineers, to be honest, because, you know, quoting and procurement we're seeing there. But yeah, we're seeing, you know, the average PO that is being sent out. The average team is spending anywhere from six to 10 hours a week monitoring that purchase order with manual time. We're cutting that down to about five to seven minutes a week. You know, we're seeing, yeah, absolutely crazy. And same with like, Judy Warner (18:34.831) Wow. Zachary Feuerstein (18:39.598) quoting a bill of materials. were seeing people take anywhere, depending on obviously the complexity and the size, up to two weeks to quote it if they're waiting on certain fabricated parts. And similarly, we brought that down to anywhere typically under 30 minutes to do, even if there are custom parts on their worth of work. those are some huge, huge time savers that we're seeing in addition to obviously saving on component costs, which is a huge Judy Warner (18:45.211) Of course. Zachary Feuerstein (19:08.714) value to our platform anywhere from, you know, call it three to 5 % plus for our average customer. So really trying to save time and money for all our customers who are out there. Judy Warner (19:20.051) I did not expect you to say that. I expected a game, but not that big. That is unbelievable almost. And I've worked, I don't know when we first met, Zach, I told you that I worked for Mill Arrow High Reliability EMS out of the DC area. And the people that I had to deal with to get those quotes done and parts and scrubbing bombs and doing all that, I was overwhelmed. I didn't know. You know, and I, as a salesperson had to wait weeks to get that stuff back and it was expected. So even though you're saying you don't have that data for engineers, I would think they'd be very motivated if all this activity is happening at their CM, that they're going to get their product back a lot faster. Zachary Feuerstein (20:11.352) Yeah, a lot faster, a lot more reliably cheaper, right? And they don't have to worry about all that back and forth that they're typically going back. can't find this part or this part's going to take way longer, or maybe you misspelled something by accident or put on like a part number. You gave only a supplier part number where the, know, the CM only had the manufacturer part number. Like we just deal with all of that so that, you know. Judy Warner (20:34.113) my gosh. Zachary Feuerstein (20:37.676) that process, you don't have to think about it. It's like buying a part on, you know, toilet paper off of Amazon, right? Electronics should be that easy. And that's what we're striving to do. Judy Warner (20:45.915) Right. That's amazing. I can't tell you how many times I personally like put the dash number on your bomb, know, and having to do that back and forth, you know, or it's like, this thing and like, just pick a capacitor. It's like, no. So I can really relate to that. That's phenomenal. Well, where can people go learn more about Zachary Feuerstein (20:54.52) Mm. Judy Warner (21:14.9) breadboard, I'm sure on your website, but is there a place they can see you in person or on your website or a webinar or a demo? You know, where can people go check it out and learn more, Zachary Feuerstein (21:27.598) Yeah, recommend everybody go to breadboard.com. From there, you can learn a little bit about us. You can also schedule a demo. I'd highly recommend seeing it live for yourself. You can even bring your own bill of materials or bills of materials and see what it would look like with your data. And we've seen even through demos sometimes people realize how much money they're wasting on their product and they're being overcharged by 10, 20, 30 % or more sometimes. Judy Warner (21:43.835) No. Zachary Feuerstein (21:55.884) and just help you with a quick win there, even if you don't work with us long-term. So, highly recommend, you know, just check us out at breadboard.com, reach out to get a demo, and we'd be happy to show you what we can do. Judy Warner (22:07.315) I love that idea of running their own bomb, right? Because where the rubber meets the road, right? In their real life world. So what a great idea. Well, Zach, I know you're a busy guy, so I really appreciate you taking time out of your busy schedule and sharing with our audience all the very cool things you're doing. And really, you and several others, I'm confident you're going to disrupt this industry. Zachary Feuerstein (22:13.633) Exactly. Judy Warner (22:31.943) but in the most positive way so engineers can spend more time making the judgment calls and doing the creative stuff that they became an engineer for. So congratulations to you and your team. Zachary Feuerstein (22:44.28) Thanks, I really appreciate it, it great talking to you. Judy Warner (22:47.037) Great talking to you. For our listeners, thanks so much for joining us today on the podcast. Be sure and go down to the show notes and get those links. I'll get them all from Zach and put them for you. And hey, go get a demo with your own bomb and see how that looks. I'd be interested in your feedback if you do go do that. So please let me know. And remember to like, subscribe and come join. We have a new website up. So come on and join our community over there at thedoubleecosystem.com. We'll see you next week. Until then, remember to always stay connected to the ecosystem. We did it. Zachary Feuerstein (23:26.74) Thank you.