Trust is good, control is better: Designing hardware faster without betting it all on AI

Stephen V. Chavez (Siemens) (00:01.27): Hi, everyone. Thanks for tuning into the Printed Circuit Podcast, where we discuss trends, challenges, and opportunities across the printed circuit engineering industry. I’m your host, Steph Chavez. As hardware has become more complex, managing details at the net level is increasingly inefficient, especially when reuse practices are inconsistent. Abstraction helps engineers work at a higher level by using reusable blocks, especially design blocks, standardized constraint templates, and reference architectures. This approach enables teams to systematically apply proven solutions instead of relying on ad hoc methods, which we know can be problematic, or individual experiences, which then becomes a single point failure. Shifting design workflows from repeatedly building systems to reusing validated design intelligence across the projects is key. This is why I’m excited to welcome back Antonio Becerra Esteban, VP of Customer Success at CELUS. Thanks for coming back onto the podcast.

Antonio Becerra (01:03.352): Thanks for having me, Steven. It’s a pleasure to be here again.

Stephen V. Chavez (Siemens) (01:07.17): Antonio, can you give our audience a brief introduction of yourself and your role at CELUS again?

Antonio Becerra (01:12.622): Absolutely, gladly. So my name is Antonio Becerra and I am a physicist turned into an electronic engineer. My career started working at Infineon, then I passed through Altium as a technical sales engineer and I landed at CELUS in 2022, if I recall properly. So a bit longer than four years ago.

Antonio Becerra (01:50.172): I started in pre-sales, also as technical sales engineer. And then over the time, I migrated to post-sales or to customer success. And now my role has been increasing to the point that I’m taking care of the whole customer success organization, namely providing, guiding the colleagues that

Antonio Becerra (02:14.342): make sure that our customers are on the supply side of our solution. The corporate manufacturers, EDA vendors, corporate distributors do get value from their partnership with CELUS and also SMB engineers on the user side, on the demand side, that they can extract value and they can really design with confidence by using AI.

Stephen V. Chavez (Siemens) (02:36.01): Awesome, awesome. With that said, let’s just jump right into it. That’s a great intro to this. So many SBUs face challenges with manual, tedious, schematic entry and fragmented workflows. Can you elaborate on real world impact of these issues on typical hardware design projects, especially for smaller teams with generalist engineers?

Antonio Becerra (02:57.164): Absolutely, that is a great point and it’s the main focus that we have at CELUS. SMBs, where I was actually talking about this a couple of months ago with Judy Warner, we recorded also a podcast, and we were talking about how thinly spread many SMBs are because they need to wear many different hats. They are under a lot of time pressure.

Antonio Becerra (03:36.854): Going from a series of requirements to a prototype, to a schematic, it requires a lot of manual and tedious work. You need to search for components, you need to read the data sheets, validate and figure out the application, find supply chain information, which means that actually from the set of requirements that you might get from marketing from a customer to the real implemented product,

Stephen V. Chavez (Siemens) (03:36.854): Mm-hmm.

Antonio Becerra (03:57.87): It’s all very fragmented, right? You need to use different tools, visit different sources of information. So this process is a bumpy road to put it somehow, which also doesn’t help when you are in a, let’s say in a small company, it might be you and your colleague or not without many resources. And you might not have a very experienced engineer.

Antonio Becerra (04:36.752): to request some support from, or you won’t have access to the FAEs from a certain company manufacturer. So you have to do many different tasks in a very time constrained timeframe and with very little resources. And nowadays in the market, it seems like the trend is there is no middle ground. You go full enterprise.

Stephen V. Chavez (Siemens) (04:56.61): Hmph.

Antonio Becerra (04:57.202): or you go full hobbyist tool. Given this landscape and how slow and tedious this effort is, CELUS fits right in to help engineers define a good architecture, system architecture. Actually, they can interact with our design assistant, our AI powered design assistant to make sure that they have a robust and reviewable system architecture that then will lead them to check compatible solutions in a way more efficient manner. Also by using our design assistant that, you know, in German, I think I might have said this last time.

Antonio Becerra (06:07.21): but it’s something that I love to use because I think it holds true, which is trust is good, control is better. That I used to say it in German, Fetraun ist good, Kontrolle ist besser. And that means, you ask the AI. I need, for example, a certain accelerometer.

Stephen V. Chavez (Siemens) (06:07.21): Hmm, nice, yes.

Antonio Becerra (06:22.382): and I need it to have on sensor machine learning. Because I’m working in a project that I need the specific solution. Which solutions are compatible with these constraints? And design assistant is gonna check the information on the solutions that we have, and it’s gonna provide you the MPNs, the applications, it’s gonna tell you where do you find on the data sheet, the confirmation, right? So it doesn’t.

Stephen V. Chavez (Siemens) (06:48.834): Hmm.

Antonio Becerra (06:49.344): it doesn’t work on the “trust me bro” basis that some AI tools work in which you just give them a very broad statement and they give you, hey, trust me what I’m giving you is what you have to go and take to production. So it really puts the engineer at the center.

Stephen V. Chavez (Siemens) (06:58.433): Mm-hmm.

Stephen V. Chavez (Siemens) (07:09.73): Mm.

Antonio Becerra (07:10.912): empowers the engineer to keep on working to provide these engineers, these generalists from small and medium sized businesses with the confidence that they are developing a robust system architecture and they are selecting the solutions that do really fit what they need at a good speed.

Stephen V. Chavez (Siemens) (07:22.626): Mm-hmm.

Antonio Becerra (07:40.67): not needing to read full data sheets just so that they can discriminate what they need. Their efficiency is boosted a lot.

Stephen V. Chavez (Siemens) (07:40.802): Mm.

Stephen V. Chavez (Siemens) (07:51.223): Yeah, I would tell you, I love it because time to market is everything. You’ve got to get your product to market and you only have a specific amount of time, your budget, resources, and you think about what it takes to get there. And the key, my takeaway of just that little, those few bullets you mentioned is how can we be or enhance the engineer to be more innovative.

Antonio Becerra (08:21.804): Absolutely.

Antonio Becerra (08:29.902): Thank you.

Antonio Becerra (08:29.902): Mm-hmm.

Antonio Becerra (08:39.278): Mm-hmm.

Stephen V. Chavez (Siemens) (08:43.07): initial design process compared to traditional methods? And what are the immediate benefits for accelerating the idea implementation cycle?

Antonio Becerra (08:51.586): That’s a great question, Steven. So let me give an example. I want to develop a Linux-based HMI system, for example. And you might be, as a generalist, you might be thinking, hmm, can I get away with a microcontroller? Probably, if you have a little bit of experience, you will figure out, I need an MPU. There’s no other way around

Stephen V. Chavez (Siemens) (09:13.622): Hmph.

Antonio Becerra (09:20.984): I’m gonna need an MPU, which means that I’m gonna need DDR memory. I might need an RTC. I’m gonna need an external flash memory probably. I’m gonna need the PMIC associated to the MPU, right? And then traditionally, you would probably go to reference designs, try to figure out a certain family of microprocessors that you know, or you have heard of.

Antonio Becerra (09:56.574): buy an evaluation board, right? Try to test functionalities. Then go to the schematic, capture everything which takes quite some time. Make sure, or try to make sure that you’re not making mistakes. And then go to the routing phase. Sweat bullets to route that DDR to route the PMIC properly to all the fun out of the MPU and everything. Or.

Stephen V. Chavez (Siemens) (10:08.354): You

Stephen V. Chavez (Siemens) (10:12.578): Mm-hmm.

Antonio Becerra (10:17.354): You could, for example, then if you work with a system architecture, you would have all these functionalities, right? You would have your input methods, would have probably a display, you have your MPU, your memories, your PMIC solution, everything interconnected in a system architecture. And it could be that you realize, my life could be much simpler if instead of using an MPU and doing everything with hammer and chisel,

Stephen V. Chavez (Siemens) (10:40.77): you

Antonio Becerra (10:46.72): I could use a SoM. I go to one of these, I explore solutions and I find out that I could also use this system on module that already contains the MPU, the PME, the DDR, many of them also wireless solutions. that means that by focusing on the high level functional blocks, on the high level system architecture,

Stephen V. Chavez (Siemens) (11:00.417): Mm.

Antonio Becerra (11:16.192): I end up having the same results, the same functionalities, which are much simpler block diagram. Because if I go for a zone that has wireless solutions, I am going to, I won’t need all these other functionalities that are all included in this one. Now, I have saved a lot of time of research. And thanks to our AI, I’m able to navigate pros and cons of doing, of…

Stephen V. Chavez (Siemens) (11:24.45): Hmm.

Antonio Becerra (11:45.336): taking path A, MPU and everything with hammer and chisel or path B of taking an SoM. Then I will know that there are functionalities that are redundant. So I won’t need to even dedicate time to figure out what to pick. And when I am picking solutions, particularly in the service design platform, one of the biggest advantages that it has is that the solutions that we expose to engineers

Antonio Becerra (12:19.83): are not only, I always bring it like that, we don’t only offer the what, but we also offer the how. So we give you the MPN, information about the whole component and the application. You are able to see a schematic, a working schematic, which means that you’re going, you’re first navigating at a very high level of abstraction. And we’re helping you as a, let’s call it like an automated FA.

Stephen V. Chavez (Siemens) (12:19.83): Mm.

Antonio Becerra (12:43.262): navigate the different complexities at a very high level. Then you can go to a lower level and find the applications that do fit and are perfectly interconnected through all the different functional blocks of your solution. And as a result, when you have validated all the solutions, you will have validated the BOM and everything, you output your schematic interconnected.

Stephen V. Chavez (Siemens) (12:46.282): Mm-hmm.

Antonio Becerra (13:09.614): Full schematic, I’m not talking just the main NPM, I’m talking all the bells and whistles, everything interconnected for you to review and then go have fun on the layout side in Siemens EDA. So that means that from the idea to the implementation, I’m going not only in record time, but also with record confidence.

Stephen V. Chavez (Siemens) (13:17.218): Hmm

Stephen V. Chavez (Siemens) (13:32.962): Mm.

Stephen V. Chavez (Siemens) (13:36.631): Mm.

Antonio Becerra (13:37.044): And the learning curve that I have experienced is much better. I have had the experience comparable to having a very experienced colleague, to having an FAE from a manufacturer company supporting me in this journey, which as someone who has developed electronics, is something that I would have wished for myself 10 years ago, to be honest.

Stephen V. Chavez (Siemens) (14:02.082): No, I would tell you it’s the evolution. And when I think about how hard it is when you first are sitting with your customer, or if you’re the customer and you’re trying to come up with how to transition the requirements into system architecture at the high level architecture, and then by the difficulty it is to translate that or get that down, broken down into a schematic level. That’s a lot of time and there’s a lot of potential for.

Antonio Becerra (14:17.41): Yes.

Antonio Becerra (14:25.454): Absolutely.

Stephen V. Chavez (Siemens) (14:28.434): Errors and I don’t mean errors as in like just simple mistakes but errors as in you’re going down the wrong path and by the time you realize it it’s too late. Whereas what you describe is a very systematic and a very much faster approach and the confidence that you described is the key is by the time you get down to it in your schematic, the confidence you have of what you created is on point. It’s exactly what you need.

Antonio Becerra (14:31.63): Absolutely. absolutely.

Antonio Becerra (14:46.241): Mm-hmm.

Stephen V. Chavez (Siemens) (14:55.254): the ability to change if need be on the fly is where you really make up time. So I love it, I love it. So let’s get into the next question I have. So how does abstraction specifically empower SMBs to overcome their resource limitations and compete more effectively, in particular when dealing with increasing design complexity? Because we know how the industry’s evolved today and the technology isn’t made for anyone. So give me your take on that.

Antonio Becerra (15:00.046): 100 %

Antonio Becerra (15:00.046): Yes.

Antonio Becerra (15:16.558): Mm-hmm.

Antonio Becerra (15:20.236): Yes. So, if you think about it, let’s put side by side, for example, big enterprises, huge teams with multiple sites, huge libraries. These companies have way bigger teams, way more resources, specialists. They might have a team, I remember in the past in another life talking to the library team, and that might have been 10 engineers across the world, only for the library.

Stephen V. Chavez (Siemens) (15:31.51): Mm-hmm.

Stephen V. Chavez (Siemens) (15:45.196): Hehe.

Antonio Becerra (15:50.062): So they have a lot of resources and they have a different way to navigate requirements. I won’t say better or worse, I will say different, right? Because then you also need to make sure that procurement has approved all the components when you’re doing something, right? So there’s different complexities. But as an SMB, you are small.

Stephen V. Chavez (Siemens) (16:09.238): Mm-hmm.

Stephen V. Chavez (Siemens) (16:14.082): Mm-hmm.

Antonio Becerra (16:16.736): as I mentioned before, right? You are wearing different hats. exactly, you’re the jack of all trades. And you have way less formalized processes, which means that you don’t have these resources of, I will ask this other colleague, or will ask, I don’t know, procurement, or I will ask my colleagues from the library to please include this and that solution. So when you work, you mentioned it before, you

Stephen V. Chavez (Siemens) (16:18.082): Mm hmm. A lot of hats, a lot of hats.

Stephen V. Chavez (Siemens) (16:28.544): Mm-hmm.

Antonio Becerra (16:46.102): you always have the risk of going down the wrong path when exploring solutions. And the worst that can happen, the worst that happens is that you’re losing time. Actually, the worst that can happen is that this goes to the market and that’s a really catastrophic situation, but let’s not go down that route. Let’s hold off there. But the key of working at this level of abstraction is that it helps you clarify.

Stephen V. Chavez (Siemens) (16:51.265): Mm-hmm.

Stephen V. Chavez (Siemens) (16:58.764): Mm-hmm.

Stephen V. Chavez (Siemens) (16:58.764): Yeah, let’s hold off there.

Antonio Becerra (17:15.116): what the requirements are and the situation in which everyone knows perfectly what is it that you need to solve, what are the requirements is, it belongs into the books, it’s not real life. So by interacting with our design assistant, you are talking, it’s like this rubber duck theorem, you’re talking to something that is helping you out.

Antonio Becerra (17:45.762): clarify resources and constraints because it can be that just because you’re writing things doesn’t mean they’re feasible. And through this conversation with this specialized assistant, you’re able to figure out that you might be asking for things that are incompatible with each other. So you are avoiding the situation in which you spend one, two weeks in a good case scenario going down a rabbit hole that actually was completely wrong.

Stephen V. Chavez (Siemens) (17:45.762): Mm.

Stephen V. Chavez (Siemens) (17:51.959): Mm-hmm.

Stephen V. Chavez (Siemens) (18:12.898): Mm-hmm.

Antonio Becerra (18:12.948): And to give you some example, in our design assistant, you can enter with natural language what you’re trying to build, and you can give as many requirements as you want, be as specific as you want. If not, the assistant will navigate through questions, trade-offs, and whatnot. I was working, I was doing a webinar last week.

Antonio Becerra (18:50.082): And I use an example, a real life example from some conversations I had, which is I want to create a hand gesture device that has an accelerometer with on sensor machine learning. And it should recognize three gestures, static hand, waving hand, and biceps curl. It should stream the data to an Android app. And the absolute main optimization factor is power saving. That was my statement.

Stephen V. Chavez (Siemens) (18:58.754): Mm-hmm.

Antonio Becerra (19:10.486): And through that, in 20 minutes, I navigated refreshing time, battery life. I navigated different constraints that I had not expressed. I set them and I found solutions that did fit them, which meant that by clarifying at a high level what I wanted to build, I managed to gain a lot of clarity. I managed to gain a lot of confidence.

Stephen V. Chavez (Siemens) (19:19.926): Hmm.

Stephen V. Chavez (Siemens) (19:19.926): Mm.

Antonio Becerra (19:38.092): and I managed to find solutions that did fit what I wanted in record time.

Stephen V. Chavez (Siemens) (19:38.882): you

Stephen V. Chavez (Siemens) (19:42.583): Yeah, you know, that really makes me think about, you know, in capturing requirements and then translating them, there are many things that get missed, you don’t think about, but then when you realize it, it increases the budget, it increases the schedule, and you have scope creep. And then now they’re wondering, why is it taking longer? Why are we over budget? And how did we get from something we thought was going to be simple to very complex and

Antonio Becerra (19:54.574): Absolutely.

Antonio Becerra (20:00.738): Mm-hmm

Antonio Becerra (20:10.093): Yes.

Stephen V. Chavez (Siemens) (20:10.816): we’re doing all these redos, what happened? And I think this approach you’re describing eliminates that. And like you said, it helps you build with confidence. That’s, yeah, that is really, really good. So let’s go to the next question I have. So how does abstraction specifically empower SMBs to overcome their resource limitations and compete more effectively, in particular when dealing with increasing design complexity? Because we know how the industry’s evolved today and the technology isn’t made for anyone. So give me your take on that.

Antonio Becerra (20:13.176): Yes.

Antonio Becerra (20:19.894): Absolutely.

Stephen V. Chavez (Siemens) (20:40.372): AI and automated schematic generation. What are some of the common misconceptions or concerns you’ve encountered and how do CELUS ensure the generated output is trustworthy? Because we mentioned this a couple of times now in this conversation and true that the content is truly usable for immediate implementation.

Antonio Becerra (20:43.47): Mm-hmm.

Antonio Becerra (20:53.218): Yes. Yes.

Antonio Becerra (21:00.302): That’s also a wonderful question that we have dealt with. We have felt that the main misconception is, and it’s actually, I wouldn’t put the responsibility on any particular user, know, or engineer, it’s the way that we’re living through now of AI, in which you are, many people are used to write a prompt,

Antonio Becerra (21:44.866): get something out and go along with it. You don’t particularly, I mean, I don’t mean to diminish anyone expertise, but if you don’t have much experience in a certain field, you’re gonna find whatever the LLM tells you, awesome, right? Great. And you’re not along with that, you’re gonna represent it. And then it could be that someone that has experience sees it and thinks, damn, like this is very shallow or doesn’t say anything.

Stephen V. Chavez (Siemens) (21:44.866): Exactly.

Antonio Becerra (21:57.102): which happens very commonly with LLMs. I have been following many posts or LinkedIn about developing hardware with AI. I saw one of an engineer whose customer asked for a certain adapter to connect two different solutions. And after getting the quote, the customer said, you know what, you’re too expensive. I’m gonna do it with ChatGPT or Claude, I don’t remember now, an LLM. And the customer went ahead and…

Antonio Becerra (22:42.502): produce those solutions. And surprise, surprise, it didn’t work. Because LLMs are wonderful to navigate information, are wonderful to compare information from different backgrounds, and are even better at boosting the capacity of those that have the expertise. the key, that’s the philosophy of CELUS, the key when working…

Stephen V. Chavez (Siemens) (23:12.543): Mm-hmm.

Antonio Becerra (23:19.746): with AI and it doesn’t need to be exclusively LLM. I’m mentioning LLMs because it’s the most common form of AI that most of the mortals are exposed to. the user, the engineer in our case has to be the one making the decisions. AI will support, AI will boost capacities, but electronic engineers.

Stephen V. Chavez (Siemens) (23:23.692): Yes.

Antonio Becerra (23:49.474): have to be the owners of every step of the hardware development, of every decision, right? I don’t want to be, I would not want to be the engineer that bets the routing. And I’m not gonna tell you on a very fancy AI, on an autorouter, right? On the autorouter without checking anything. And I go and produce it. If that thing works, if that thing works without me having…

Antonio Becerra (23:49.474): better tune the autorouter, given better constraints, review the routing, God knows what would happen. So the key is that, you give your prompt, you give your statement to the design assistant, and the design assistant is gonna generate something for you to review. You can then change it by hand, can ask then the design assistant to change it for you. You can ask, hey, you have added this, you have added that. I was thinking more in this direction, this other direction.

Stephen V. Chavez (Siemens) (24:06.934): Mm-hmm.

Antonio Becerra (24:20.494): It can’t help you, but it should not make decisions for you. And to talk about trust, for example, we have been very vocal lately about our collaboration with NXP. NXP, their block diagram system, is now powered by CELUS, is an instance of our tool.

Stephen V. Chavez (Siemens) (24:28.258): Hmm.

Antonio Becerra (24:50.322): The solutions that Siemens users, for example, see on the Siemens instance coming from manufacturers, they are developed in collaboration with manufacturers. So that means that you, as an engineer, are going to use design assistant to help you navigate the constraints, generate the system architecture, you making the decisions. And then when you’re gonna check the solutions, those solutions are generated

Stephen V. Chavez (Siemens) (24:58.977): Mm-hmm.

Antonio Becerra (25:20.82): in collaboration with the component manufacturers with a human in the loop. I actually, one of the teams that I am leading is in Porto and is responsible for this, which means that engineers that have this fear of the unknown, fear of I’m going to ask something and God knows if this will work, with CELUS it’s not a trust me bro, I need to build this and go along with whatever the platform produced.

Stephen V. Chavez (Siemens) (25:26.551): Mm.

Stephen V. Chavez (Siemens) (25:45.237): Yeah.

Antonio Becerra (25:47.682): you are on the driving seat, you are making the decisions. And it won’t take long. I did it in 20 minutes in a webinar. So any concern about efficiency, any concern about I’m actually gonna need longer than if I did it in the way that I’m used to, it dissipated very, very quickly. I had an example on LinkedIn, a user that started writing that the platform hallucinated and…

Stephen V. Chavez (Siemens) (26:08.706): Mm-hmm.

Stephen V. Chavez (Siemens) (26:17.474): Hmm.

Antonio Becerra (26:17.638): and that it generated something that it wasn’t what he wanted. And we had a very nice session in which we explained precisely this. Hey, you have to take into account that there is a series, if you don’t interact, if you don’t own the decisions, then I wouldn’t trust any output from any system. I mentioned just the autorouter, which is, know, old technology.

Stephen V. Chavez (Siemens) (26:44.482): Mm-hmm.

Antonio Becerra (26:44.494): So you have to be the owner, you have to make decisions. And I think it was one session in which we give a quick onboarding to the platform. And he was so kind to actually even post a couple of weeks later, another piece on LinkedIn, giving wonderful feedback, valuable feedback, a part positive and a part actually something to take in to improve the product. But the trust had changed.

Stephen V. Chavez (Siemens) (27:00.733): Hmm?

Stephen V. Chavez (Siemens) (27:09.857): Mm-hmm.

Antonio Becerra (27:14.604): He was now trusting the tool because he was making the decisions and he saw how CELUS addresses this topic and how much value he got in a very short timeframe.

Stephen V. Chavez (Siemens) (27:15.01): Mmm.

Stephen V. Chavez (Siemens) (27:28.652): You know, I really love your comment that it sticks to me is, you know, the engineer has to own the decision. I mean, that is key because utilizing the tools at hand or that at your fingertips, you know, it’s a culture shift, you know, a culture shift of the fear of the unknown, concerns about efficiency, trust in automation, you know, and how do we apply this? And engineers that embrace this evolution

Antonio Becerra (27:57.07): Absolutely.

Stephen V. Chavez (Siemens) (28:24.642): will flourish, those that don’t, they’re gonna get left behind. And there’s many times that I’m speaking in conferences and I’ll ask the question to the audience, how many of you use automation? How many of you embrace this? And you’ll be surprised, very few hands will come up. A lot of people are afraid of the unknown and they stay within that pocket of where they’re used to being and stepping outside that. And those engineers that do embrace it and step outside their comfort zone,

Antonio Becerra (28:31.63): percent.

Antonio Becerra (28:35.822): You

Antonio Becerra (28:42.382): It’s not. It’s not. Yes. Exactly. The engineer that knows how to use AI is going to take your job. Not the AI. Indeed. That’s that’s.

Stephen V. Chavez (Siemens) (28:52.938): Yeah, exactly. let’s go into the next one. When we think about the evolution, looking beyond the immediate benefits, how does working with abstraction and tools like CELUS prepare engineers and companies for future trends in electronic design, such as even greater automation and the increasing complexity of hardware systems?

Antonio Becerra (28:56.718): Mm-hmm.

Antonio Becerra (29:15.992): So I think precisely when we talk about complexity is when engineers, I think the value that one can extract of tool like CELUS and the CELUS design platform increases massively with complexity. But for the same topic we were discussing 10 minutes ago, which is you are an engineer in a team of three or in team of two, or you are on your own, right?

Stephen V. Chavez (Siemens) (29:32.672): Hmm. I agree.

Antonio Becerra (29:46.292): and you’ve been asked, hey, I have this project, can you take it? And as time goes by, projects won’t get any simpler, right? Things are every time smaller, more power hungry, they have to be more efficient. So it is not like over time, the requirements and the requests from customers are gonna get any easier. So being able to adopt a tool like CELUS.

Stephen V. Chavez (Siemens) (29:56.898): No they won’t.

Antonio Becerra (30:15.962): Knowing how to work with a trustworthy AI system that acts as an experienced engineer, acts as an FA, is what is going to help engineers give the next leap in quality, the next leap in confidence, as I’ve mentioned before, and be able to take on more and more complex tasks, complex designs.

Antonio Becerra (30:53.41): as I mentioned before, and Linux-based HMI system, but it can be a point of sale, right? It can be some home automation. Maybe you are an engineer that for the last five years, you were happy with STM32 microcontrollers, which are wonderful. You’re used to their ecosystem, now you have to create a point of sale, Linux based point of sale.

Stephen V. Chavez (Siemens) (30:53.41): Mm-hmm.

Antonio Becerra (31:14.926): What do you do now? Damn, right? Without automation, without the power of an electronics focused AI, the amount of time that you’re gonna need and how error prone this project could be, it can make the difference between succeeding or failing and having to do something else in life. So, right? You go…

Stephen V. Chavez (Siemens) (31:34.21): Mm.

Stephen V. Chavez (Siemens) (31:40.994): Yeah, you’re absolutely right.

Antonio Becerra (31:44.672): You check the platform, you give your statement, you navigate your solutions, and maybe you find out that you need to use an IMX95 from NXP or an SoM from Variscite or Toradex to make it easier to connect with what I was saying before about focusing on the architecture. And this type of support that you would only get at a huge company at an enterprise level.

Antonio Becerra (32:14.882): from an FAE, from the component manufacturer or from a very experienced colleague, small companies cannot get it unless they embrace automation boosting their capacity.

Stephen V. Chavez (Siemens) (32:14.882): Mm-hmm.

Stephen V. Chavez (Siemens) (32:24.866): Yeah, I mean, the small and medium business as well as individual on their own cannot afford to miss the window of opportunity. That window, that marketing window, they cannot afford to miss it. And whereas the large enterprises, yeah, they could be late and the hit is not as significant as if you’re a small to medium business. I mean, there’s a chance you miss that window of opportunity and your company folds. It’s that simple. It’s that simple. You know what?

Antonio Becerra (32:34.862): Mm-hmm.

Antonio Becerra (32:49.3): Exactly. Exactly.

Stephen V. Chavez (Siemens) (32:53.374): I can, you and I can keep going on this topic, you know, hours and hours, but you know, we’re running out of time here. So with that said, you know, that’s about all the time we have for today. Is there anything else you’d like to add,

Antonio Becerra (32:55.918): Yes, yes,

Antonio Becerra (33:05.294): Just something very quick because we’ve been talking a lot about SMBs and for any SMBs listening to the podcast, we have launched the CELUS Success program. So if you’re starting to check the platform on the Siemens instance, on Siemens Connect or any other instance, just reach out, send an email to sales at celus.io. You can introduce yourself, your company and if you want…

Antonio Becerra (33:43.892): the project that you’re working with or the work you need your next challenge. And we will give you high touch onboarding. We will help you out. We’ll teach you how to work with the platform, how to empower you as an engineer, to make sure that you are confident in the decisions that you’re making. And we can take it from there.

Stephen V. Chavez (Siemens) (33:56.131): Awesome, awesome. And for those, if you’re not following Antonio or myself on LinkedIn, please we do. Please, I hope you follow us as well as CELUS and Siemens. That’s a lot of great content we’re sharing with the industry. So with that said, Antonio, I want to thank you for sharing your insights on abstraction. To our audience, check out our latest or my latest blog.

Stephen V. Chavez (Siemens) (34:33.27): titled from manual effort to engineering intelligence. We’re driving PCB productivity through automation and abstraction. I also want to recommend checking out the webinar by David Haboud titled accelerated PCB design productivity with AI and intelligent automation. As always, I want to continue to have you or continue for you to tune in and for, let me say that again.

Stephen V. Chavez (Siemens) (34:52.118): As always, continue to tune in for more trends, challenges, and opportunities across the printed circuit engineering industry.

Antonio Becerra (34:52.118): It was great to be here, Stephen. See you next time.