Hello, good morning, good afternoon, depending on where you're joining us from. I am Mike Genovese, the Communications Equipment Analyst at Rosenblatt Securities. This is the Age of AI Tech Conference, and as always, I'm super happy to be joined by Calix, one of my favorite stocks, one of the best stories in the industry. We have Carl Russo, the Chairman of the company, and Cory Sindelar, the CFO. Good to see you, gentlemen. Good to see you, Mike. Great. We did this in June. We're going to do it again now, and I'm interested to see the incremental differences or similarities between what we spoke about in June and what we spoke about now. Let's begin just at a high level, Carl and Cory, just a quick overview of the company's mission and strategy and help position Calix for those in the audience that are newer to the story. We prefer Calix, but Calix, whatever, just remember it's not Calix, that's all we do know. Look, the mission hasn't changed in the 15+ years we've been pursuing the vision of a service provider space that was going to be disrupted by what we called broadband service providers at the time. That legacy wireline and cable and wireless and enterprise networks would disappear, and they would be replaced by a broadband network with a broadband service provider. That broadband service provider would take advantage of the data and the metadata that was coursing through their networks and use it to deliver an unparalleled subscriber experience. We call those folks broadband experience providers. That story remains constant, literally for the last 15 years, and everything you've seen us build is literally a stepwise, layered approach to a platform model with our platform-based OSes, our cloud-managed services, and obviously heading into the future. As you've heard Michael speak to on the earnings call, we are heading into an agentic AI world. It is the same and continuing to build from there. Right. I just want to let the audience know, as we have this discussion and questions come up, please type your questions into the Zoom interface, and they'll come right to me. I will make sure to ask the questions to the Calix team. Team, let's talk about the confidence that you've been exuding. You're exuding it now. You were exuding it last quarter. You were exuding it the quarter before. I'm simply talking about confidence in the business outlook. Where does it come from? We're obviously in a higher rate environment than a couple of years ago. We have less new broadband experience providers being formed by the private equity market. What is it that existing customers are doing that gives you visibility and allows you to have the confidence you seem to have? It's about that, back to that disruption. It's hastening, and it's beginning to cross the chasm, as you've heard us speaking to. In essence, it's the difference between a home's past and a subscriber's served world. When you have low interest rates and a lot of competitors that are running on incapable physical infrastructure, if you build fiber, you'll get a certain take rate. As interest rates go from basically negative real interest rates to now positive real interest rates, to your point, when they move in between here and here, there's a lot of uncertainty, but they've been stable for quite a while. You've seen people settle in and say, okay, I know what the interest rates can be. Now what are they focusing on? The competitive environment continues to get stronger, and it's really about the subscriber experience. It hasn't changed for us. The demand is driven by the number of subscribers that are coming onto the model via our customers, the ARPUs that are growing, the churn rates that are declining, the Net Promoter Scores that are going up, and we just focus every day on helping our customers succeed. The visibility actually continues to get strong and stronger. The appliances during the pandemic went through its own wave, and as a matter of fact, there's a very nice three-minute video of Cory on our investor website that helps you understand what actually happened. The software and services side has continued to grow every day, unabated. Okay, great. I guess from here, I mean, since I've been following you, and for longer than I've been following you, the revenue growth has been roughly 15% average, maybe even slightly higher, with about 1.75 points of gross margin improvement per year, except during the pandemic, which I think Cory talks about on that video that you described. As we sit here today, is that still the right way to think about the company, mid-teens growth and almost 200 basis points of gross margin improvement per year? It is Cory's video, so I'm going to let him speak to it. Yeah, Mike, nothing has changed from our long-term model, which states that we'll grow somewhere, you know, 10%- 15% per annum on the top line, that we'll see 100- 200 basis points of margin expansion. We don't see that changing anytime soon. We can go through the OpEx. From a From a perspective, that's what we see. We still think we can operate within those ranges. Okay, I mean, one thing that investors have been asking me about is that if you look at the last two quarters, the sequential growth in 1Q, I think was 7%. In 2Q, it was even higher at 10%. I think some of that had to do with taking share from DZS, a small competitor that went bankrupt. I'm constantly getting this question about, you know, does quarter-to-quarter growth have to be slower going forward? Obviously, if you grew 10% quarter- over- quarter every quarter, or even 7%, you would blow away 15% top line. I think that we probably shouldn't expect that much growth. Just help me answer my clients who ask me that question, which is, you know, can Calix keep up this type of sequential growth? How should I frame that for them? As a very large shareholder of the company, I would love to see that. I think the way I would couch it, and Cory, certainly feel free to add color, is the continuous commitment to sequential growth is actually, even at a low number, a big challenge, having grown sequentially 7% and 10%. The whole model has shifted up. You ask a good point. There is clearly a footprint that is becoming available to us to talk to customers about. The whole business model, when that door cracks open, when you have an event, like a smaller competitor, a file that goes directly to chapter seven. Let's keep in mind, it's not like they had billions of dollars of footprint, but they did have a few hundred million dollars of footprint. That does afford you the opportunity to increase your bookings as you go out, win the customer, get an order, etc. The surprise was not that we were going to win a lot of these customers. We expected to win a lot of them and have them cap what they had done and grow from there. A number of them decided not only to cap and grow, but actually to rip out and replace a lot of what they had purchased. Once they made those decisions, that caused them to want to get started quickly. That was the unanticipated shipments that drove a large part of that revenue increase in the quarter. Just to be clear, you're talking about just with this one bankrupt competitor that you've seen them ripping out equipment or these other competitors? Yeah, you don't. Look, there's also a footprint available to us in legacy accounts that may have had a different vendor. As you go forward and you enter that account, they're typically not going to rip and replace what's already there. To be blunt, we likely would not counsel them to do it if it's working and moving the bits. We'd rather them invest those dollars in someplace where they can have a material effect on subscriber experience. There's that as well. It's the, I got to get going now on a rip and replace that drives that upside. Cory, feel free. No, I agree. I echo that. A few answers ago, Carl, you were talking about the importance of subscribers to your model, your customer subscribers, which I think, you know, we're talking about subscribers for which Calix would be receiving, you know, recurring revenue from, which makes them your subscribers too. We're sitting here and we're, as analysts, expecting you guys to begin to talk about appliance revenue and appliance margins versus, you know, cloud and software. I don't know if I'm breaking that out with the correct terms. If you want to correct that, versus, you know, cloud and software revenue and margins. We're really looking forward to that info. I think that'll be great. You've also said that you'll probably, you know, are very reluctant to ever share your subscriber data and give us a number on subscribers or for ASPs. I respect that, but I just want to be reminded of sort of what the reason is, why that is so sensitive. Because on one hand, it would help people analyze the company so much more, but on the other hand, you know, if you're being this successful doing what you're doing, we don't want you to change. Just help explain that for us. I love it when people say, please don't ever change, Carl. Let's try this on the first piece. We have said for a long time that breaking appliances out from the gross revenue stream would allow competitors to better foot the model. Be happy to take if there's a trading discount that people believe. We'll take that trading discount rather than foment more competitors. Cory has convinced me that as the gap between us and potential legacy competitors continues to widen, that reason sort of falls on false logic. That's why between now and the end of next year, we will break out the appliance revenue. Having said that, I don't think we will ever give a subscriber count for a number of reasons, which is now you're starting to talk about math that you're footing it to the last cell on the spreadsheet. You're also starting to give out data that relates to your customers, even in aggregate. That's not ours to give in that way. Anything that relates to subscriber data, remember, we have real-time telemetry coming in from the networks and real-time information about their subscribers. We treat that data with the highest security. It is sacrosanct. Anything relating to it, we simply won't give out, and a subscriber count would be one of those things. Okay, that's fair. I think that we maybe back up a little bit here because we talked about your growth being sort of mid-teens or higher with seeing very impressive gross margin improvement. We've talked about in the past, I think that it looks like the margins over time are headed at least into the high 60s. For those who are sitting in the audience kind of scratching their head and saying, I just didn't think the broadband market was that strong. I just didn't think the broadband market would support that type of growth. I don't see Nokia doing that. I don't see Adtran doing that. This doesn't make sense to me. Help me answer the question of how a company like Calix can grow mid-teens with gross margin improvement every year in a market that people think is probably growing 1/3 of that rate or, if that. Great question. Because you look at the market and you say, you know, this has been around for years. It's a mature market. Actually, it's not. That's the point of the disruption. The legacy markets are mature markets. The coaxial HFC, the twisted pair of copper, the wireless markets are very mature markets. What is not mature are the broadband experience providers that are actually disrupting the market. That's the immature part of the market that's growing rapidly. It's not us in a mature market hacking out, you know, 2x or 3x or 4x the growth rate of the market. The only way you would be doing that would be, to your point, driving your margins into the floor in an undifferentiated environment. It's not so much what we are doing that is differentiated. It is, but it's the platform that we enable our customers to become that's differentiated. Our customers are differentiated to their subscribers, and they are in an immature market because they are disrupting the subscriber experience market. That's what makes what appears to be a mature market an immature market. Let me give you a simple parallel. I had a cell phone, then a flip phone comes along. Then the iPhone comes along. Are those the same markets or are they different markets? At one level, they're all in the same large market, but one is quite disruptive from a technology standpoint. The difference in our world, same level of disruption, but these markets last for a long weekend. We're in the infrastructure space, and that's a multiple decades-long disruption. It's hard to see, but that's what's driving it. Hopefully, that makes sense. I just, you know, like where I live, I don't think that I have a provider that would have an MPI score over 10 or, you know. I just don't think that's available to me, right? I feel like a lot of my clients are also coming from places where we just, maybe we don't have the Calix power b y the way, that speaks to literally our opportunity. The vast majority of subscribers are unaddressed by what we do. As we cross the chasm and have ever larger customers, all of a sudden, we start to reach more and more subscribers with the model as those service providers become more successful as they transform into experience providers. That is why the phase of crossing the chasm, where you go from the early adopters into the early majority, is where we are. We are hearing the cell bells starting to turn up, and you hear it in the names from the customers that we talk about or customers that we do not talk about yet. You did talk about a new tier one customer on last quarter's call, and part of that was a large carrier acquired a customer of yours. I think that same customer seemed to have made a separate decision to invest in your cloud technology. Can you talk about that? What I heard Michael and Cory say was a large cloud customer. In our parlance, a large is $2.5 million subscribers or larger, which basically makes it a Tier 1. Purchased a cloud, in essence, independently of everything else. I didn't say it, they did. Sorry. That's where it's left. I'll give you a couple of historical behaviors. We don't announce anything unless it's up and running, ever. You will never hear Calix, and if you go back through time and memoriam, we've never talked about lab trials or this thing or that thing. If it isn't in, up and running, and the customer's happy, we would never talk about it. This was merely to help folks understand that this platform is crossing the chasm. You are now able to, in essence, bring a large customer on board with simply our cloud. That's a milestone. As you know, seven years ago, Verizon came on board with our AXOS E-Series systems. Those are very different sales to sort of different people. Okay. I think one of the highlights from the quarter was the announcement of what you guys characterize as the third-generation platform with agentic AI. On one hand, when we hear third-generation platform, it makes it seem like it's an entirely new appliance, entirely new software. On the other hand, I understand this is adding an agentic AI sort of umbrella framework to the existing architecture, and it's not maybe as big of a backend change for you as some people might think it is. Can you help us understand, on one hand, going from non-agentic AI platform two to agentic AI and platform three is a really big deal. From a sort of an investment that Calix has to do, how big of a deal is this new platform? Let me see if I can, if you'll indulge, I'm actually going to start at square one and help you understand how it fits into our evolution and our customer's evolution so that it puts it into the proper framework for what it is, what it isn't. When we began this journey on this vision of a broadband service provider becoming a broadband experience provider, we built AXOS and EXOS as the abstracted operating systems that run the appliances in the network and on the premises. We brought to market a cloud. At the time, we called it Compass Cloud in April of 2011. In essence, that brought in all of the real-time telemetry and started to convert it into offerings. Those offerings were oriented around tools. In 2016, we started, actually, this was what Michael brought to the company, actually, when I was interviewing him. He said, you know, if you take those tools and orient them this way to the personas that you're using them, you can unlock an enormous amount of value for the customer and for the company. Big light bulb went on. We, in fact, re-architected our clouds to literally do this, to become persona-based, and all the tools were subordinated inside of those personas. If you were a service person, you still had a saw and pliers, but they were a different saw and pliers than what the marketing person might have had. That was, in essence, our 2.0 version of the platform. On top of that, we started to add managed services, which then got bundled into SmartHome and SmartBiz and SmartTown, and most recently, SmartMDU, which is in its early phases of launching. That's the go-to-market that you saw. That's what's growing. That's what's continuing to enable our customers to win subscribers and be successful. Now, what of this platform 3.0 and agentic AI? Let me break them out. Forget the agentic AI for a moment. What's core and central to the platform 3.0? There's really a couple of things. Platform 3.0 allows our clouds to do two things that they did not do before. One is they can function in a private cloud. Today, our cloud is in our cloud. If you were a service provider and you wanted to run that cloud on-prem in your cloud, could not do it, would not do it. That is a big enhancement. We expect, as we continue to cross the chasm, that we're going to have larger service providers that want to run it in their cloud on-prem. Not an unusual thing. The second thing it allows us to do is to place the cloud in sovereign data centers. As you know, if you go around the world, there are lots of different regulations on personal information, personal data, data residency, etc. We believe, for the most part, as we start to look to expand internationally, we will be advantaged by being able to place our cloud instance in a sovereign data center in that country. It enables us to expand into geographies and customer segments that we might not otherwise reach with our platform 2.0 cloud. Let me stop there and see if that helps you understand what's going on with that 3.0 cloud notion. Before I go to agentic AI. Yes, from the cloud perspective, that helps. Okay, so you've seen us continue to expand product-market segments. SmartMDU is an expansion of a product-market segment. By the way, to talk about expansions, SmartMDU is actually the largest expansion that we've done in some time. Why? Because fully 1/3 of people live in multiple dwelling units, which, if you do the math, if we're addressing 2/3 and we're adding 1/3, it's actually a 50% increase in us being able to address the subscriber marketplace. It's a huge step up for us. Those sorts of expansions continue to go on. Now, what are the agents? The agents run, in essence, on top of the model. They benefit from all the telemetry, from the data, from the operating systems into the cloud, right into the managed services. You can see how that all works. The agents are there, first and foremost, to help our customers go faster. It is fundamentally a constraint for all of our broadband experience providers that they can go faster than they are staffed to do in almost every circumstance. These agents will literally leverage what they have and enable them to get more subscribers, by the way, help those subscribers purchase additional services that are valuable to them, and on and on and on. The number one goal of the agent is to help our customers go faster. Let me stop there and see if that helps. It helps. My next question that kind of leads to is, when do you know if, when will you know if the agents are successfully helping your customers go faster? This just launched. When will we have data to know whether this is working? When will you have it or when will we have it? Trust me, we'll have it and we may never give it to you, but that's a separate issue. I mean, we'll start seeing it late this year, early next year. Ultimately, the way it will manifest itself is the software and services will grow faster. That's where it manifests itself. Will there be a fee, all those sorts of things? Sure. That's not what's going to drive the software and services. It's the agent actually speeding up the deployment of what we already know is a proven winner for our customers. It sounds like there should be a revenue benefit coming. I guess my question is, is the OpEx or investment or the CapEx investment you needed to make in these agentic AI capabilities, is it already in the model or is it an expense that has to come before the revenues accelerate? A lot of it is in past tense and was in the model. I mean, will the model change? What can change the model? Two things. Revenue performance, right? Revenue performance, gross margin expansion, as you know, allows the model to have a few more dollars. The other thing that you've seen us do from time to time, and I'm not prognosticating this, I'm really saying the following. If we see an opportunity to forward invest and speed those curves, we will do that. It would be premature. Right now we're just head down executing. As you saw, let me give you an example. When the pandemic curve, Cory's video happened, we made the decision to hold OpEx flat, not take OpEx down to maintain the OpEx model. The reason we chose to do that was, in our minds, we saw a great opportunity to go land new footprint and grow. By the way, that decision from a year and a half ago yielded the first and second quarter of this year. I mean, that's what drove that outsized sequential growth because if we had cut OpEx, I would guarantee you would not have seen that kind of burn. If there's an opportunity, we think, to go faster with effect, not go faster for the sake of going faster, I would be remiss if I didn't say the team may make that decision. As we sit here today, an enormous amount of this investment is behind us. Cory, please feel free to. Yeah, Mike, I'll spend a moment covering the accounting. As we were developing the next cloud, a portion of that is being capitalized, right? That'll then come back and go through co-ops for a period of time until that initial development effort gets amortized. Call it a three to five-year window, that'll go away. However, I'd also say we're also, as we transition service providers, cloud providers, we're not as efficient as we were in the past. There's a little bit of an overhang. You can kind of think of it as mostly a wash, right? For example, as we go from one to another, ordinarily you would go out and buy three-year agreements and buy down some of your costing. You can't go do that. You're getting a little bit more inefficiency today. Whatever benefits we're getting from the capitalization piece of it, it's kind of offset by lesser cloud costs, if that makes sense. What happens. I think when you have more experience with this, you'll be able to plan out three years in advance because it's so new, you can't go and do it. No, no. As you're transitioning off of one vendor to another vendor, you don't enter into those long-term contracts with the old vendor. You've done that with the new vendor, right? You've got that all laid out, and you've got that, in a sense, a little bit of, you know, yeah, it's running too. You have an overlap time between the two clouds. You have an overlap time between 2.0 and 3.0. You're running dual until you're up and running. That is also what Cory's referring to. Yeah, even within this conversation, I mean, from an overall company level, for the next year or, you know, year and a half, I don't know, there's more, from my perspective, right? There's more operating leverage in the model than we're used to seeing, right? We're used to seeing you being at model and trying to invest in revenue growth. Now, I mean, we're talking about these issues, but these issues seem less important because we actually have a good deal of leverage in the model for at least the next year. That's correct. Thinking about that, right? That's correct. You are thinking about that correctly. Yep. Anything to talk about on macro or tariffs, just kind of important headlines? How to the way on that? Sorry, just kidding. Cory, you want to spend a moment on, you know, just what you're doing there, balance sheet, supply chain, what you're seeing, Jerry, etc.? Sure. Everything I say is predicated on what we know today. As you know, that could change tomorrow. It could change this afternoon. I haven't gotten to the news tonight, so maybe it happens already. That being said, we are fortunate that the tariffs that have been implemented to date don't affect us given the equipment that we have that we're importing under, right? They've been exempted under that. Again, stroke of a pen, that could change. We hope to keep our head low and try to avoid any kind of distraction that that may result in if that got changed. We're not standing still on that. We understand it's a very uncertain environment. We are working towards a strategy where we have capacity both in Southeast Asia as well as here in Mexico, on the North American continent. There's still a price differential to manufacture in Mexico versus Southeast Asia. There will always be kind of a cost-benefit of what is the actual tariff rate that would get implied and whether we would move things. Nevertheless, we want to have a presence in both places to give us complete optionality to determine if that were to come into play. For the moment, things have been stable. Where we do see tariffs today in the kind of perverted way is when we're building America, buying America. You might have seen some tariffs on imports of components into the United States. There's a talk about tariffing semiconductors coming into the States if your supplier doesn't have manufacturing in the United States. There's a lot of stuff moving. That'll increase costs as it relates to manufacturing our Buy America, Build America components. That's about the only place where we see it at the moment, knock on wood. Yeah, so I would just add one piece. There are small tariff impacts that we have absorbed, but they're small enough in the grand scheme of our model where you haven't seen them affect anything. We've done nothing on pricing actions, and we'll not do anything on pricing actions unless and until we see something where it's very clear that this has to be done. The team has the balance sheet and a narrow SKU set, which is really important. When you have a small number of SKUs, you're going to be willing to invest out into the future to grab components and other things. If you have a high number of SKUs, you're always fearing excess and obsolete in the future. I think comparatively, we're in a wonderful position. Great. Carl, when I asked you earlier, I tried to ask a question about crossing the chasm to larger service providers, but I focused it too much on the news of this [OneCloud] win from last quarter. Just a more general question about crossing the chasm and how things are playing out in terms of more medium and large customers being interested in your solutions, what you're seeing from that perspective. Lots of conversations, lots of interest. Interest is necessary but not sufficient. Pretty clearly, if you don't have it, you might as well hang it up. That interest continues to expand at a very good rate. These things take time to work their way through the funnel. It is very clear from where the conversations, at what layers the conversations are having in our large, medium, and small, what the future will indicate. It's coming. Okay. What about separation from competitors? Who is your closest comp, and is there a relevant grouping of how we should look at the competitive players out there? Actually, I was thinking about that earlier when you were talking about margin expansion into the high 60s. I was thinking back, look, when we were in the 30s, getting a four in front of it was exciting. When we were in the 40s, getting a five in front of it was exciting. We're in the 50s, obviously now approaching 60, getting a six in front of it is exciting. Why do I say that as the reposed answer to your question? One of the challenges we have and one of the challenges that investors have is there isn't a comp to what we're doing. I'm not sure how to answer it. There isn't a comp. Are there substitutes inside of a legacy service provider to do things in a different way? Of course. Our model is built around the broadband experience provider and making sure that we're doing our best to help people come along. The answer is no, not really. Okay. Continuing with this theme of re-asking questions a little bit more focused on earlier questions that I asked. Even before you announced agentic AI, I thought about your cloud business as being a business that was in some way sub-scale, meaning that the incremental margins are probably higher than the total margins. This is an opportunity, right, to reach scale and to improve margins. I think this is a big, you know, cloud was probably a big driver. Now agentic is part of that. Is it a similar, like, should we be thinking about, you know, cloud being sub-scale and needing to, you know, and needing to reach and driving margins? Basically, agentic is just a continuation of that story. It's not that different than before, or is there something that I'm missing that makes agentic different? I'll let Cory go first. Yeah, Mike. Part of the re-architecting of our clouds is obviously taking the benefit of everything we learned over the six months and coming up with a new architecture. Part of the design goal was to help improve our margins, right? To take in scalability into that and take those learnings and transform it. You're right. Part of the new architecture helps us drive that, right? Inside of our improving gross margins, cloud margins will improve themselves over the next period of time. Irrespective of the COGs and the margin of the agent itself, the fact that it is speeding the ladder just simply drives more leverage into the clouds, i.e., it puts more transactions on the clouds, which are coming in at a higher incremental margin than the average margin for the cloud. We have lots of levers to continue our gross margin expansion. Yeah, just that the timing of seeing agentic AI accelerate RPO growth, what should we be looking for? Next year, you should not, not in the near term. Okay. Like Michael, and I would say, and to Carl's earlier point, the best way to think about the agents is they make our customers go faster. The real benefit is the acceleration of the model being pulled forward as more subscribers come onto the network, and you're monetizing AXOS, EXOS, the clouds, and the managed services, right? You're monetizing the platform. It's not so much the agent. In order for that to show up in RPOs, somebody has to negotiate a new agreement. Otherwise, they're just paying incremental software fees that you don't see in RPOs. I just want to make a last call. We have five minutes left. Just a last call to the audience to submit questions through the Zoom interface so that I can ask the questions to the management team here. If no other questions come in, I can easily fill five minutes. Obviously, it sounds like SmartMDU is extremely important. It is. You characterize it as sort of adding 50% to the subscriber TAM is what I heard. If you want to double- click on that and talk about that anymore, and/or if there's any other new cloud-managed service offerings that we should be tracking that are particularly important. None of the latter. On the SmartMDU, it's an example of the power of the platform that we've built because it enables our service providers to literally deliver that broadband experience to an MDU as an extension of the platform they're already running. It is literally shake and bake. If you go into multi-dwelling units, hospitality units, it's a very fragmented market with all sorts of downsized enterprise solutions sort of wedged into the application, upsized home solutions which step on each other. It hasn't really been engineered to do that. It's the same problem that exists in small businesses. Small businesses have never been designed for. They've always either been take an enterprise solution and downsize it or try to scale up a residential solution, neither of which gets to the real need of a small business owner. It literally shows it's a showcase for the power of the whole platform that we built. It enables our customers to literally go and grab that subscriber share at very low incremental costs for them, if that makes sense. They're not retraining all of their field technicians or re-engineering their whole network. They've got an entirely different go-to-market selling approach. It all fits straight into what they're doing. Now they're adding incremental markets at very high incremental margins. What I've found is when you can help your customers add revenues at very high margins, they're happy and they're happy to share a bit of it with you. Yes. Okay, great. In the few minutes that we have left, let's talk about BEAD. I'd like you to kind of touch one more time on how important or not BEAD is to the Calix investment thesis. Specifically, I saw that Virginia and Louisiana have submitted their BEAD applications. Can you help us understand that news and if that's meaningful for the company? The answer is it's meaningful directionally because, as you may remember, when BEAD new administration came in, there was lots of storm and drawing around low-earth orbiting satellites and what's going to happen here and fixed wireless and the benefit of the bargain, etc., etc., etc. At the time, we said first, it's not in our model, by the way. That is true today. Not in the model. However, what you're seeing come in from Louisiana and Virginia, Virginia first and Louisiana second, are actually interesting indicators of where this may land, which is you're seeing the soft dollars that were in the original BEAD awards because there were roughly 1/3 of the dollars in there for training and other things as opposed to building infrastructure. Actually, what you're seeing is those States saying, we'll give that back to you. On the stuff we're actually going to use to build infrastructure, literally 1% or 2% is going to low-earth orbiting satellites. The largest percent is going to integrated fiber and fixed wireless applicants. You might assume we know a lot of them, and I would leave it at that. In actuality, after all of this churning, it looks like it's coming back to it's going to be mostly fiber with fixed wireless endpoints. Guess what? All the states that we're hearing, say very few, want to have a scalable solution. If they're going to invest federal monies that are going to have to last them for a decade or more, they want the solution to be able to scale over that Internet growth as opposed to something like a low-earth orbiting satellite that actually doesn't scale. That's what we're seeing. All right, that's promising. All good news and not in the model. Thank you. We've run out of time. We could keep going, but we've run out of time. I actually have to jump on a hurry's call now. Carl and Cory, thanks so much for joining us. It's good to see you coming on here, continuing to exhibit confidence. We'll expect continued sequential growth out of you guys. Maybe it doesn't have to be 7%- 10%, but. Thank you for that. We will continue to put one foot in front of the other and stay tuned. Thank you, Mike. Perfect. Thank you for everybody for joining us in the audience as well. I'll talk to you soon.