largest players in machine vision and one of the earliest innovators in industrial AI, which of course is Cognex. Joining us from the company is the company's Chief Financial Officer, Dennis Fehr. Dennis, thanks for being with us today. Ken, thanks for having me. Yeah, of course. So, maybe just to start, for anybody who might be a little unfamiliar with Cognex, could you just give us a real quick, brief overview of what the company does, where it sits in the industrial supply chain, and what you're seeing in the market today? Great. So, maybe first, quick intro on myself, Dennis Fehr, Chief Financial Officer. About two and a half years with the company. So, Cognex, as already said, we're about 45 years into machine vision, probably the category creator, and we see ourselves as the tech leader, especially also leading with the latest AI machine vision tools. We use machine vision in factories and in warehouses to inspect, think about defect inspection. We do identification, like barcode reading and what we call optical character recognition. We do also robotic arm guidance and measurements, and we do that in about a $7 billion market, which is growing about a 10% to 11% CAGR, as somebody like Interact Analysis would say. We are serving that market largely through a direct sales force, serving some of the most iconic names in manufacturing and in warehouse and logistics. We recently also started to revitalize our channel program to serve system integrators and machine builders better. We are a high-margin business, so our long-term average is about 28% adjusted EBITDA margin, and for this year, we are guiding to 29%-31%, so we are also in a nice period of margin expansion. Great. That is a really great intro and a great overview. Maybe just to start off on my end, you reported earnings last week. You put up a really strong beat and raise. You introduced a new full year 2026 guide range. The stock did sell off on the day, and I think part of the comments I had gotten back was nominally, it does look like the fourth quarter guide implies, or it implies the fourth quarter earnings kind of steps down from a 3Q level. Curious if you could just talk through that a little bit. What are the puts and takes on how you think about the visibility into the end of the year relative to the strength that you have been seeing in the end markets? Great. Maybe I will start with a bit of a macro picture, and then maybe go from there some more of the quarterly sequential view there. Big picture macro, right? We upgraded four of our five end market in terms of our growth outlook, and we see really a strong demand environment. If we look at PMIs, we are now about six months into PMI being in expansion territory, and they are around 55. That means if you think six months just in, if you look at past cycles, you would have seen PMIs peaking somewhere in the low 60s, maybe in the mid-60s. In that regard, we feel like in general on the cycle, there is still quite some way to go, right? We still feel like we are in the early stages of the up cycles, and in general, we feel quite robust about the macro environment. We think that we see some of that reflected in the numbers which we put out there. We reported a second quarter with record quarterly revenues. For the first time in this $290 million range, and then we put out a third quarter guide with a midpoint of $310. That means another quarterly record, if we make that number. In that regard, we see two strong quarters, and these are two quarters, however, we need to point out, are quarters which are driven by electronics seasonality. Typically, electronics is showing up with strong numbers in the second and the third quarter, and electronics does not have a meaningful impact in Q1 and Q4. Electronics is the largest growing end market if we take absolute dollars. In that regard, clearly, it is a growth driver in second and third quarter, but it is not reflective of driving growth in the fourth quarter just by seasonality. If you look at the fourth quarter, it is also a number we put out there which was slightly ahead of consensus before we went into the earnings season, and certainly now it does not have the electronics contribution from growth, but it is still a growing quarter. In that regard, we would say we would not take this as an indication that the growth rate year-over-year and that there is a sequential step down from Q3 into Q4 as any signs of a weaker demand environment. We think it is more a seasonality effect. Understood. That is really helpful. Dennis Fehr, you and the CEO, Matt Moschner, I think has made a really impressive impact on streamlining the cost structure over the last, call it, two plus years. First, you are expecting to realize about $35 million of cost saves annualized by the end of this year. You have also are exiting around $22 million of low growth, low margin business over the next year as well. You have raised the long-term EBITDA margin target by about 300 basis points through the cycle. Obviously, there is still a lot of work to be done on what has been announced, but I am curious, how do you think about the further opportunities to optimize the cost structure from here as we execute against the current initiatives? Yeah, no. First of all, very pleased about the results, right? If you look back, 2024 was a year where we just were in the high teens of adjusted EBITDA margin, 17% to be precise. Now in this year, we are guiding to the 29% to 31% for this year. In that regard, over a comparatively short period of time, we achieved quite some margin expansion. What were the ingredients for that? On the one side, very clearly returning back to the growth, the top-line growth. For this year or for the first half of the year, somewhere in the mid-teens of growth. At the same time, the ability to take out OpEx. You mentioned we have this $35 million OpEx reduction target, and we really see this, that 2026 total absolute adjusted OpEx numbers will be below the 2025 absolute numbers. That means we get very, very strong leverage. In the second quarter, we had 100% flow-through from revenue to the bottom line. For the full year, we said that probably in the high 80s, 87%, probably flow-through of revenue into the bottom line. Now as we start to look forward, where would we go from here? We mentioned that we are probably good with resetting the cost base, so we are not looking to take out further cost in 2027. At the same time, we also do not see the need that we need to add significantly more cost as the top line further grows. In that regard, we can further drive automation initiatives, we can drive additional process improvement initiatives. That basically sets us up that we could keep an OpEx growth somewhere at an inflationary level, and that would mean we would still see a strong margin top-line flow-through to the bottom line, probably not at an 87% basis, but you would think like in a long-term steady state, maybe 60%, and then 2027 could be somewhere between the 60% and this 87%. In that regard, still strong flow-through, and then certainly depends on what the top line will do in 2027. Maybe too early to talk about that yet. But in general, are we at the end of our margin expansion story? No, clearly, we are not. I think we clearly have that ambition to go further. Right. To that point, that long-term target of, what is it? I want to say 25% to 31%. Yes. You did 32% EBITDA margin this last quarter. The midpoint of the third quarter guide is 33%. I guess, is there a ceiling on where EBITDA margins can go in terms of through cycle performance? When we think about a 25% to 31%, we think about this as an annual number. Certainly, quarters like the second and the third quarter where we have strong electronic seasonality, they may be outside of that range. But then for the full year, our guide of 29% to 31%, the midpoint would be 30%. That's still well within the range. And certainly, we will look then further. In that regard, for us right now, the focus is about let's achieve in the next five months what we put out there. And then once we have the full year behind us, we will reevaluate and we'll basically come out whether we will change that range or not. Understood. Maybe just switching over real quickly to the end markets. You talked a little bit about consumer electronics. It is about 20% of your total revenue. Despite, I think historically, a decent amount of that exposure had been in China, but it feels like you have seen more upside from that sector as it has moved out of China and into broader Asia and even Europe on some of the supply chain moves. Even with all the moving pieces and the inflation, what has been the biggest driver of growth in that vertical and where do you think that could settle out through a cycle? Yeah, no, the great thing of consumer electronics in this year is that it is a very broad-based growth. That means if we look back at prior cycles that were typically very focused on certain things like a change of technology, like a display technology, if we go back to 2017, for example. Right now we see it very broad based. That means we have factors like reallocation of supply chains. That means things moving out of China. That was probably a bigger theme last year than it is this year, but it is still helping this year. We have new form factors. We have new devices overall. That means you see a different, completely new form factors. That means not a different phone form factor, but think about glasses and other wearables, which are in the making and in part already launched. Then we see data centers as well, bringing a complete new component into what we so far called consumer electronics end market. Which per se is not consumer electronics. It is really a complete new set of an end market to some extent, which we have not served in the past. In that regard, what we see, it is a very diversified growth. In that regard, that keeps us actually very positive and very confident that this would not be like just a one-year growth in consumer electronics or electronics broader, and that we could see that extending for a bit longer period. Yeah. In that regard, I would say, we feel really more positive about this end market than probably we did in prior cycles. Yeah. I guess just to touch on the form factor change, right? You do have a large customer within that sector. There has been a lot more news about them potentially introducing a foldable phone later this year. How do you think about the opportunity from that? You have got some experience through other suppliers or manufacturers on foldables. I am just curious how you think about how big of a catalyst that could be. I would say at the moment, I would more point back to the broad-based growth. That means there are many different factors which drive growth, and form factors of phones may be one of them, but we are not thinking about it like this is the one and the big thing which really drives. As mentioned before, we are thinking at the moment really about how strong the growth is from a broad-based perspective. And again, we see it as actually quite positive. Yep. That makes sense. Maybe switching over to another end market that has been showing some solid growth, semiconductors is about 10% of your portfolio today, right? Right. It seems the market is finally seeing a re-acceleration in orders, and obviously supply chain tightness has kind of exacerbated that to a certain extent. Right. You want to talk a little bit about what you are seeing from the sales perspective on semis versus maybe the challenges internally that you are trying to overcome in terms of delivering on higher supply chain challenges? Sure. Absolutely. I think very well known to everyone, there's a huge demand out there, especially on the memory side. That drives, first and foremost, very outsized demand for our products. As we serve really the broad spectrum of machine builders in the semi space. So, our customers would be the semi CapEx names out there. We are very well specced in into their machines. So, in that regard, we feel like we can really grow with this market, as this market is expanding with pretty good certainty there. It's really the market which has the highest growth rate in terms of percentage in this year, and that's kind of what we, to some extent, also expected. When we came into the year, we were not sure about the timing, but we were sure that it would be happening for all the good reasons I mentioned just before. So, in that regard, it's driving a strong demand. It's driving strong growth for us, and that's really overwhelmingly, obviously, a positive thing for us. But at the same time, we also use memory chips in our products. However, it's not, let's say, a major factor in our bill of material, but since we are seeing really more like 2x to 3x of price increases over a 2025 baseline. Certainly, it starts to show up into the P&L. Originally, we said it might be 50 basis points in the third quarter, and we upped it up to 75 basis points. But at the same time, we also very clearly are confident that we are offsetting that with our own pricing actions. So, in that regard, we see it as a temporary effect in the P&L. Then again, overall, I want to say that overwhelmingly, we are a net beneficiary from that trend because what it does to our top line and what it does to margin accretion to our business, we clearly think that let's say, this memory side is actually a good thing for Cognex. Right. Makes sense. Maybe before I continue, I just want to open it up real quickly to the floor in case there's any questions. If not, then we can always continue. Any questions so far? All right. That's fine. We'll continue. I've got a long list here. Look, I've continued to get questions from investors on whether AI development is an opportunity or competitive risk to you guys' longer term. I view you as one of the first true innovators in industrial AI, just being, like you said, you kind of created the market of machine vision. You've been a public company since the 90s, right? How do you kind of respond to that process of, is the democratization of AI and people being able to make their own models, how do you view that as a potential risk longer term, or is it a catalyst for you? We think it's much more an opportunity and a catalyst than it's a risk. As you already rightfully said, we created the category, and then about 10 years ago, and that's almost how long we're already in our AI journey, we really started to embrace AI and saw what it may mean as a potential catalyst for bringing machine vision inspection task onto the factory floor. In late 2017, early 2018, we acquired a company in Switzerland called ViDi, and they really kind of formed the basis for our AI development. Then we launched our first AI-enabled product in 2022. Since then, each additional product which we have been launching has had some form of AI machine vision to it. In that regard, it's really a technology which we have embraced a lot, and it brings two opportunities. It brings an opportunity of penetration, and it brings the opportunity of adoption. Let me unpack that a little bit more. On the penetration side, it really enables to do machine vision inspection task, which will not be possible with prior rules-based machine vision tools. We just showcased in our recent earnings call, the data center full rack assembly. That wouldn't have been possible a year ago. That was only be recently possible by launching our deep learning OneVision Cloud platform, and in combination with a new machine vision system on the edge, the In-Sight 3900, which is powered by Qualcomm. That means you now have the training modules on the cloud, and you have the power and the speed of the newly launched edge device, and only that combination actually made it possible to introduce such a full rack for server rack inspections. In that regard, it creates market. In that regard, this step into the data center supply chain is, for us, a market creation step, which was not there. We think this is incrementally positive, and that's the penetration aspect. Then there's the adoption piece, is that it's very clearly that machine vision is not something easy to use. We try to make it as easy to use as we can, but if you want to do a harder inspection task, then it comes with challenges to set it up and so on, and here AI can also help. A simple example is that last year we launched an AI assistant which basically helps with auto-config and auto-setup in some of the devices which we have. That's not available yet in all abilities of our portfolio, but that's basically a starting point in that journey as well. That means bringing AI and making it easier for customers to adopt it, so that they don't need to have big factory automation engineering teams by themself. But somebody with a, maybe like you and me one day, can set up a machine vision system for a medium to complex machine vision task, and that's what we're striving to do, and that's where AI can help. Yep. We have a question here. Go ahead. Yeah, just following up on that, are you seeing multimodal LLMs encroaching into the traditional machine learning vision space where people are trying to use LLMs instead of actually training a vision model for their tasks? Not really. Certainly you can say, is there a risk? The problem is still that machine vision for factory automation is highly specific. You really need to identify small defects, and then it's very specific. That means you really need to train. That's what we do. We pre-train models with highly specific factory automation related data. In that regard, while these large language models are getting larger and larger, they're also trying to do more and more things. That means they're not really equipped to solve the most specific factory automation tasks. In that regard, we don't perceive them as a risk, and therefore, to your initial question, we think the opportunity is so much larger than the risk. Thank you. Sure. To that point, I think something that we've talked about in the past is because you are a first mover into this industry, you've got decades of high-fidelity image libraries for industrial widgets that you can't train an LLM off the internet to get with high accuracy or precision. Is that Exactly. Yeah. Yeah, you nailed it down. Yep. Maybe just sticking on the AI side, you introduced OneVision at your analyst day last year. I was really happy to see that you've made it commercially available at some of these past trade shows that we've seen you at this past year. Maybe talk about the trends and what is OneVision for people who don't know, and how do you think about attach rates associated with that service and what it really opens up for you? Right. OneVision really extends our ecosystem to enable basically running the most complex inspection task on a device. That means if you think about if you want to do a complex machine vision task like the server rack assembly verification, then you need to run them in a deep learning mode to train, and you probably have to label a lot of inspection points you want to do. That compute power you will not have on a device. But at the end, you want to do the inspection on a device because you don't want to have the cybersecurity implications, and you may have latency issues if you go to the cloud. What we created with OneVision is basically an online training platform where the customers can upload their proprietary data, train their models, then redeploy the model from the cloud into the edge. We call that an edge-to-cloud and cloud-to-edge strategy, giving the customers what they need. They need the cloud training capabilities, but they want to have the speed of the edge and the data security of the edge. We give them both with OneVision. That basically is what makes it so exciting, is that we really are the first one going out with such kind of an offering to the market. While we just recently launched it basically two months ago and making it fully commercially available, we see good attach rates, and we are pretty pleased where we are in this even so short journey. Yeah. We have about a minute left. Just want to open it up for any final questions for Dennis. Okay, maybe just last one from me. You do have a target of around 300 basis points of inorganic growth through the cycle. When I typically think about your M&A track record, it is typically been for companies that are pre-revenue and getting you into the door on a new type of technology vertical, something to scale on. It seems like that has shifted more towards companies that are truly already generating revenue and more in the software side. Maybe talk a little bit about what you are seeing out there and just how quickly you feel like you could go after some of those opportunities. Right. In general, I think we are pretty pleased with the technology stack which we have, so we feel pretty confident about our AI capabilities, about the ecosystem which we have been building. That means we do not feel like we need to do a lot of tech bolt-on acquisitions. At the same time, we have been talking at the last earnings call about the opportunity to further drive growth through diversification, and that is where M&A could come into play and to look at entering additional adjacencies where we could either leverage our existing technology, where there would be technology synergies or where there would be sales general synergies. That is the opportunity which we have, so in that regard, that is an opportunity, and certainly nothing to announce at the moment, but we will keep you posted, and we are definitely looking in this area to make potential acquisitions. Perfect. Well, I do think that is us at time, but I really want to appreciate everybody for their time and thanks to you, Dennis.