Operator
Good afternoon, and welcome to Duos Technologies' third quarter 2026 earnings presentation. I'd like to remind everyone that this call is being recorded. Earlier this afternoon, the company released its third quarter 2026 financial results via press release and filed its Form 10-Q for the three months ended September 30, 2026. Joining us for today's call are Duo's CEO, Doug Recker, and CFO, Chris D'Almeda, and following management's remarks, we will open the call to questions from the company's covering analysts. Individual investors who have additional questions not addressed through live Q&A may contact Duo's investor relations team at duot at duotech.com. Before we conclude today's call, I'll provide the necessary cautions regarding the forwarding statements made by management during this call. With that, I'll turn it over to Doug. Sir, you may proceed.
Thank you. Welcome, and thank you all for joining us this afternoon. I'm going to spend most of my time today on where DUIS is going, not on walking through the quarter line by line. Chris will take you through the numbers, but this quarter's numbers are a snapshot of the company in the middle of change. The more useful conversation is about who we are, where we're going, and what we're building. So here are the headlines. In the third quarter, we completed the transformation of Duos into a co-location compute center company. We develop, own, and operate compute centers built on our modular architectural design. We're targeting sites from 1 to 20 meg across the United States, with a driving focus on sites between 5 and 10 megawatt going forward. We provide the building, the power, the cooling, the connectivity, and the security. Our customers bring the compute. That's the business now. During the year, we've executed three major transformations, starting with selling our legacy rail business. We also wound down the asset management work for the new APR Energy, and on September 30th, we sold our GPU as a service entity to Axe Compute. Every dollar of capital, every hour of management time now goes to one thing, compute centers. I want to spend a moment on the last step, because it's the deliberate choice we made. Our GPU business did its job. It proved we could find sites, bring power online, and stand up high-density AI infrastructure for demanding customers. But owning GPUs is a different business from owning the compute center. GPUs can lose value and carry technology risk. Compute Community centers give us long-term contracts and assets that hold their value. With the GPU-as-a-Service SPV sale that we announced at the end of the quarter, Axe covered approximately $87.8 million of debt tied to the GPU fleet. And ACT will pay us an additional $42.9 million in monthly installments of $715,000, with the remaining balance due no later than September 30, 2027, or earlier if ACT or an affiliate obtains financing secured by the GPU cluster. As for Duos, we continue to serve the end customer from our Columbus campus just as before, but now under an extended five-year agreement, we've entered into a lease of the GPUs with acts to service the customer. Additionally, as our balance sheet no longer carries GP risk, our capital now goes to compute centers development. So why compute centers, and why focus on the 5 to 10 megawatt range? The demand for high-performance computing keeps growing, but the bottleneck has moved. It's no longer chips. It's power and it's time. Large hyperscale campuses can take years to permit, connect to the grid, and build. Meanwhile, cloud providers, AI platforms, and enterprise need capacity now. The capacity is beyond everyday life. This shows you in the stream at night, on the online game your kids play, The AI tool that students use to study for exams, and the telehealth visits with doctors and other cities. More and more, the capacity needs to be spread across more locations closer to the people and businesses using it. That's the gap we're built for. A 5 to 10 megawatt size site is big enough to matter to a cloud provider, yet small enough that we are able to find stranded power without waiting years in utility queue. That opens up markets to large developers overlooking, include Tier 3 and Tier 4 markets. And because our architecture is modular, we can bring capacity online in months, not years, and add more in phases as additional customers sign. Just as important, we build differently. Our compute centers use closed-loop cooling so they don't consume the millions of gallons of water many large facilities need. We go where the power already exists. Open capacity that's sitting unused. That's stranded power, so we're not adding strain to the grid that local families and businesses depend on. And we aim to leave every community better than we found it. Better connectivity in areas that have been underserved, local jobs and engineering, construction and operations, and new tax revenue for schools, roads, Public services and local infrastructure. In many of the towns we operate, this makes a real difference. Columbus, Georgia is the proof. We acquired the campus in July of this year. Within about 75 days, we had installed more than 7 megawatts of new infrastructure, all with local workforce, and the campus began full operation in October. Today, it supports a cluster of 2,304 NVIDIA B300 GPUs, and we have a five-year, five-megawatt agreement now. V2Power is what our customers are paying for, and Columbus shows we can deliver it. It also is our template. What we did in Columbus is what we intend to repeat in every site. The second proof point... It's who's signing with us. Last month, we signed a 10-year, 20-megawatt co-location agreement with a mature, publicly traded cloud service provider. It's valued at more than $400 million in total contract value, and deployment is expected to begin in the second quarter of 2027. That agreement tells you three things. A 10-year term is twice as long as our earlier agreements. It came from a sophisticated counterparty that did its due diligence and chose DUOS. And the 20 megawatt is exactly the size of site we're built to deliver. So where are we headed? Let me be specific about the next 12 months. First, power and land. We're working to secure approximately 100 megawatt of additional power to compute center zone land across multiple U.S. markets. About 25 megawatt of that is already under contract, and we're in active negotiations on the other 75. Second, delivery. Just as we did in Columbus, we're already developing the 20 megawatt site for our new cloud customer, which is scheduled to go live in the second quarter of 2027. And we're on track to deliver at or before the ready-for-service date. We're also developing an additional site in Georgia. They could go live as early as the first quarter of 2027. Stay tuned for more information on that in the future. Additionally, we're seeing stronger demand from larger, better finance, and more diversified customer base than we did six months ago. This is a strong indication of how far we have come in such a short period of time, and we are headed. Third, funding. We plan to fund our growth primarily with non-dilutive financing at both the project level and the corporate level, match the contracts we've signed. Chris will say more about that in a few minutes. Additionally, let me update you where we are with the other agreements we have announced. We are still progressing with the 10-megawatt site announced earlier this year, but working through the additional power delivery from Georgia Power has delayed the project. However, our equipment is already on site, and we have a solution that we expect will be up in Q2 of 2027. Additionally, we are still progressing on the 55 megawatts and actively looking at sites to support that capacity. Finally, our technology solutions group is growing rapidly. Duos technology solutions became our largest revenue line this year, and the amazing team is not stopping anytime soon. This group serves customers building data center and AI infrastructure across the country. It also gives us sourcing scale and supply chain assets for our own compute centers built. When Longleague Equipment decides who delivers on time, that's a real advantage. Before I turn it over to Chris, I want to thank our team. What we've done this year is remarkable. In less than 12 months, this team has sold our legacy rail business, wound down the new APR agreement, Raised more than $120 million, acquired and stood up at Columbus campus, completed the sale of the GPU business, and signed a 10-year agreement with a publicly traded cloud service provider. Any one of those would be huge, big, big year for the company our size. His team did all of them while running the business every day. I especially want to recognize our field and construction crews in Columbus. They installed more than seven megawatts in about 75 days, and that's the kind of execution our customers need. I also want to thank our technology solutions team who built our largest revenue line from zero in about a year, and thank you to our board and our shareholders who backed us through a year of big decisions. We've also strengthened our leadership team this next phase. During that past few months, we've welcomed Dipan Patel, our new Chief Operating Officer, who's leading the build-out of our site, as well as Matt Paolim, who joined September as our Chief Legal Officer. And also, I'm glad to have Chris Guilameda, our new CFO, Chris actually helping to guide the capital rate decisions and focus on funding the business with minimal equity solution. With that, I'll turn it over to Chris.
Chris? Thanks, Doug, and thanks again for joining us. Since this is my first call with Duo, a quick introduction. I joined as CFO in late August after spending most of my career as a CFO in the publicly traded construction and infrastructure industry. My priorities are simple. Give investors a clear, consistent view of how this business makes money. Fund our growth in a way that protects shareholders and build a financial discipline and reporting structure a company at our stage needs. Because the business has changed so much this quarter, let me start with three things that will help you read the results. First, the rail business, which we sold on August 5th, is presented as discontinued operations. Second, this quarter shows the business in transition. The asset management revenue that drove the third quarter of last year has wound down. Technology Solutions is now our largest revenue line, and recurring co-location compute center revenue is just starting, with the Columbus site having begun full operations in October after the quarter closed. Third, the quarter includes several items tied to the GPU sale, including a gain on the sale and a loss on paying off the related bets. Now, beginning with our top line, total revenue for the third quarter of 2026 was $10.5 million, up over 100% from $5.2 million in the third quarter of 2025. Technology solutions contributed $6.8 million, compared with virtually nothing a year ago. Services and consulting revenue was $3.5 million, which was primarily related to work for a new co-location customer. Related party services and consulting revenue for the third quarter of 2026 was approximately $70,000, down from $5.2 million in the prior year period as the asset management agreement wound down. Gross profit for the third quarter of 2026 was $4.4 million, up 150% as compared to the same period a year ago. This led to a gross margin of 42% for the third quarter of 2026 as compared to A 33% gross margin for the prior year period. This improvement came from a shift in the revenue mix away from the asset management work. Technology solutions earned a margin of about 15%, which is typical for a sourcing and integration business. The professional services work carried a much higher margin. Over time, we expect recurring co-location revenue to become the main driver of our margin, which will continue to drive gross margin improvement as more megawatts come online. Operating expense for the third quarter of 2026 was $6.2 million, compared with $2.6 million a year ago. The increase came mainly from higher, non-cash, stock-based compensation, along with investment in business development and the team we need to scale. Stock-based compensation for the third quarter of 2026 was $2.3 million. This resulted in a third quarter 2026 loss from operations of $1.9 million, compared with a loss of operations of $900,000 in the prior year period. Other income for the third quarter of 2026 was $1.4 million. This included a $4.7 million gain on the deconsolidation of the GPU as a service business and a $740,000 of interest income. These were partially offset by a $2.8 million loss on extinguishment of debt and a $1.6 million of interest expense, both tied to the GPU facility that actually paid at closing. The loss from continuing operations before income taxes was $500,000, compared with a loss of $800,000 in the prior year period. Net income for the third quarter of 2026 was $864,000, or $0.02 per fully diluted share, as compared with a net loss of $1,000,000, or a loss of $0.06 per fully diluted share in the prior year period. Adjusted EBITDA for the third quarter of 2026 was approximately $500,000. The company did not report adjusted EBITDA in the prior year period. For the first nine months of 2026, revenue was $18.8 million, up 36% from the prior year period, with a gross profit of $8.9 million for the first nine months of 2026, or an increase of 87% from the prior year period. Income from continuing operations before tax was $50.1 million, which includes a one-time $53.5 million gain during the second quarter from realizing our interest in the parent of new APR energy. Excluding this gain, loss from continuing operations before income tax for the first nine months of 2026 was $3.4 million, compared with a loss of $3.3 million in the prior year period. We ended the quarter with $44.8 million in cash, compared with $15.5 million at the end of 2025. During the quarter, we spent significant cash on ordering long lead items for current and future projects, as well as securing additional sites for development and the finalization of the Columbus build-out. As of September 30, 2026, we had essentially zero debt outstanding. We remain comfortable that cash flow from operations and current receivables will continue to support the ongoing operations of the business outside of future site development. Now, for anyone modeling the company, here's how our revenue works. We track each site in three stages, signed, in-service, and billing. Recurring revenue starts when a customer's capacity is ready and accepted, not when the contract is signed. That's why recurring co-location revenue starts to materially show in the fourth quarter as the Columbus site has a full quarter of operations. Several of you have asked us to be clear on what a site earns, so let me take this opportunity to clarify the details for the framework going forward. Now, I realize some of these numbers are updated from what we've said previously, but this is a business that has moved quickly and refined its focus and operating model in real time. As a result, we have streamlined our approach to the markets and reduced operating costs for future sites. Moving forward, our total CAPEX built per megawatt, including land and site work, is approximately 6.5 million. This is the per megawatt CAPEX level we can expect on each site for the total available megawatts per site, not just the critical IT load. Now let's look at each megawatt from a return perspective. At our current contract rates, each billable megawatt generates roughly two to two and a half million of revenue per year. A billable megawatt is what we can contractually bill the customer for. Typically, this is the critical IT load. On the cost side, we expect to pass power through to the customer going forward, eliminating power costs for the critical IT load from our operating site expenses. Therefore, after site operating costs, each megawatt should produce around $1.5 million of site-level EBITDA on an annual basis. I also want to clear up the payback question. Some of our earlier materials referred to a payback of about less than three years. That figure compared revenue to build cost. Measured on a site-level EBITDA, excluding financing costs but including non-critical IT power costs which we would pay for, the payback is just under five years. With project debt or lower non-critical IT power costs, the payback on our own equity is considerably shorter. However, returns will vary by site. The biggest drivers are the contract rate, the cost of power, and how much additional work a site needs to be fully operational. That's why we pick sites with stranded power first and sign the customer before we build. Now, as we enter the fourth quarter, the transformation of Duos into a co-location compute center company is materially complete. Our operating model is now built around long-term, recurring contracts, plus the addition of our rapidly growing technology solutions infrastructure business. Based on contracts already performing, a full quarter of co-location revenues, and orders scheduled for delivery in the fourth quarter, we expect fourth quarter 2026 revenue to be at least $30 million. As a result, we remain comfortable that total revenue for 2026 will exceed $50 million, and we expect positive adjusted EBITDA for the full year. I want to point out that the sale of the GPUs as a service transaction will impact EBITDA going forward, including the fourth quarter. Specifically, we are no longer paying interest on the GPU financing, but now we are paying GPU rent that runs through cost of revenues. This, of course, moves the expense from an EBITDA add-back to a site-specific operating expense. As a result, selling the GPUs changes the EBITDA calculation and will affect Q4, but, as I just said, we still expect positive adjusted EBITDA for the full year 2026. Regarding the full year 2027 guidance, I have been here a little less than three months and still trying to fully clarify all the puts and takes in a very fast-changing business. Additionally, I always want to provide relevant, clear guidance so investors can understand where we are and what we are looking to achieve. In that regard, we are not changing the previously stated $160 million of full-year 2027 revenues. However, given some of the recent changes in the business, it is premature to provide a full-year 2027 EBITDA range. In any case, we anticipate we will see solid EBITDA growth in 2027 and plan to give full-year guidance on our year-end 2026 earnings call. Now, with the transformation of the business material complete, our near-term priorities are clear. We'll finalize site agreements on approximately 100 megawatts of powered land, and we'll prepare for commencement of the 20 megawatt cloud deployments in the second quarter of 2027. With that, let's talk about the question we hear most from shareholders. How are we going to pay for this? Starting with the size of it. At roughly 6.5 million per megawatt, our next 75 to 100 megawatts is a program of roughly 480 to 650 million built over several years. That's a big number for a company our size, but we won't build it all at once. We'll build it site by site and in phases, and we'll continue to commit construction capital to a site once a customer has signed. Here's how we will fund each site going forward. Our goal is to fund a majority of each site using non-dilutive, project-level, debt and debt-like instruments secured by site and its customer contracts. Long-term contracts with credit-worthy customers, like our 10-year cloud service agreement, are exactly what lenders want to finance. The rest of the site build is equity in the project, and we have several sources for it, including cash on our balance sheet, receivables that are on our balance sheet today, continuing cash flow from operating sites, and, where it makes sense, partners who invest alongside of us at the project level. Corporate level financing sits on top of that and gives us the flexibility across the portfolio. So what does this capital cost? Well, I won't quote rates before terms are final, but what I can tell you is how we think about returns. A site must earn a yield well above what it costs to finance it. With unlevered returns in the mid-20s, 20% levels, that spread gives us room for today's debt rate levels while still creating value for shareholders. On dilution, our clear preference is not to issue common equity to fund the business. That said, we are evaluating a range of structures, and we won't rule out any tool, but every financing decision gets the same test. Does it grow megawatts and recurring cash flow per share? Equity remains one of the tools available to us, and if we were to use it, it would be to fund growth with returns we believe are clearly accreted to shareholders over time. But let me be clear, our number one focus and top priority is non-dilutive financing at the project and corporate level. With that said, here's where things stand today. Columbus has been funded from our balance sheet. For the 20-megawatt cloud deployment and our next sites, we're in active discussions with several potential lenders and capital providers at both the project and corporate level, and we'll update the market as agreements are finalized. However, we are not stopping. We have already begun work securing sites and ordering of equipment for upcoming contracts. This is reflected in the cash we used during the third quarter, and that is solely focused on driving site development and increased future EBITDA. Like I said earlier, our cash and expected ongoing cash flow from operations will cover our current non-site related commitments. External financing is what lets us build the next sites at the pace our customers want. To sum it up, the business model is now straightforward. We sign the customer, build in phases, and finance each site against its contracted cash flow with a clear preference for non-dilutive capital. We'll measure our success by megawatts in service and recurring cash flow per share, and we'll keep reporting our progress on that basis. Going forward, my job is to make sure every dollar we put to work earns its keep for shareholders. With that, I'll turn the call back over to Doug. Thanks, Chris.
Let me leave you with three things. One, Duelist is now a co-location compute center company. The model is simple. We build, run the compute center, and our customers bring the compute. Two, the model works. In Columbus, we installed more than 7 megawatts in about 75 days, and the campus is full operation. Additionally, a publicly traded cloud service provider just chose to partner with us for a 10-year deal. Why? Because they've seen what we can deliver. Three, the runway in front of us, about 100 megawatt of powered land and development, focused on distributed sites from 5 to 10 megawatt in markets and communities that we partner with that others overlook, and each one funded site-by-site against signed contracts. That's key. We're building something durable, and we appreciate your support. Operator, would you like to open it up for questions, please?
Operator
Thank you. The company will now be taking questions from its covering analysts. If you would like to ask a question, please press star 1 on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star 2 to remove yourself from the queue. For participants using speaker equipment, it may be necessary to pick up the handset before pressing these star keys. Now, our first question will come from Scott Buck with Titan Partners. Please proceed with your question.
Hi. Good afternoon, guys, and I appreciate all the added color over the prepared commentary, especially around the financing. First question, I'm curious, Doug, the 20-megawatt CSP contract, is the 2Q start a full 20 megawatts, or is that a phased ramp?
...megawatt on when it's delivered, whether the customer is installed or not.
Okay. So, if you deliver a full 20 by the second quarter, they're paying for a full 20?
That's correct. Yeah. Okay. Perfect. And then that CSP deployment, that's on the existing flagship campus or that's on a new site?
I'm sorry, you broke up. So, the 10 meg, is that what you're asking, the additional 10 meg?
No, no, the 20 meg, the 20 meg contract.
No, no, no, no. The 20 meg is not in Columbus. It is not. It's a new facility.
Okay. All right, perfect. And then I guess my last question on the land pipeline with 100 megawatt land pipeline. Said you have about 25 megawatts currently under contract and 75 megawatts in negotiation. What is the timeline to close there? And then, are any of these sites tied to specific customer demand, or is this somewhat of a speculative build?
Sure, the 75 megawatt, actually most of it is under LOI and hard money down, so deposits are down. I think we should know a majority of that by the end of this year, so going into January, but those sites, there are multiple sites, obviously, average amount of 10 to 20 meg per site. Those are going to be picked based on customer worthiness, basically, credit worthiness. So we're in a position now where we can pick and choose. Our backlog is huge. And, you know, six, eight months ago, it would have been a different story. But now we're looking for the future. We're looking for longer term deals and high credit worthy customers.
And one thing to keep in mind, Scott, is we don't do speculative. So we look for signed contracts with customers before we build.
Perfect. That's it for me, guys. I appreciate the time. Thank you.
Operator
Thank you. Thank you, sir.
Operator
Thank you. And as a reminder, if anyone has any questions, you may press star 1 on your telephone keypad to join the queue. Our next question comes from the line of Ed Wu with Ascendient Capital Markets. Please proceed with your question.
Ed Wu
Analyst — Ascending Capital Markets
Yeah, congratulations on all the progress. You guys are very close to, obviously, what's going on with all the data centers and AI trends. Are you seeing any big changes so far, particularly, obviously, with the recent political change in political landscape?
Yeah, actually, great question. So our secret sauce right now is obviously, if you follow us for the last year, you see where we deploy in Tier 3 and Tier 4 markets, right? So we've actually had the pitch down pretty well. We go into these markets first, and we actually sell the community first before we actually engage in the land, because what we don't want to do is basically buy the land or put money down on the land and find out the community does not want us. So every piece of property that we're talking about today and in the future, we've already vetted the community. Their buy-in is there. There is one site where I have a final buy-in. I believe that's mid-December in Ohio, and that's one of our larger sites. But yeah, the key here is to educate the market. When we go into a market right now, the first thing you're getting is we don't want data centers, and we understand that. We totally understand that. So what we do, and DuPont has been excellent at this in our marketing team, is they go into the market and they do basically a show and tell. So they'll go into the community center and they'll have booths set up where the community can come in and we show them what we're building. We also show them what a mega campus looks like because that's what they think everybody's building when you say data center. So when they actually see our layout and how we build and we don't use water and we're in the community, we're actually there to help them grow their connectivity and everything. They actually like it. And what's very interesting, what's really cool about what we're doing is the majority of our sites, I would say nine out of 10 of them, they come to our Columbus facility. So just last week, we had one of the towns come with their chamber of commerce with their elected officials. And they came in and toured the facility. We've had a couple other folks do that as well. And we welcome that because they can see the size of the facility we're building that's totally different. They actually bring, you know, they calculate the noise. They look at the water intake. So it's not we're just saying it. They're actually, we're willing, hey, come look at it. Come see what we're doing. And we actually educate the community. So that's worked out huge for us. And plus, we're going to where the power is already, right? So they're not building new infrastructure out in the town. We're not taking and using power that the community is already using. It's stranded power, basically, is what we call it. So it's at the transformer level, and it's sitting there at the substation. That's our approach.
Ed Wu
Analyst — Ascending Capital Markets
Great. That's a great answer to my questions. Thank you, and I wish you guys continued good luck. Thank you. Thank you, sir.
Operator
Thank you. Our next question comes from the line of Bill Papa-Nestaggio with Charity Capital Markets. Please proceed with your question.
Yeah, good evening. Thank you, gentlemen, for taking my questions. Yes, sir. Yeah, great. Just wanted to add on to the prior question. Doug was hoping he'd be able to share a boots-on-the-ground perspective of the supply-demand environment for data center colo capacity and how you expect it to evolve. You know, I'm curious to hear how. Availability of powered land opportunities are trending and whether you're seeing market dynamic shift given the regulatory headwinds. Do you anticipate new entrants will start seeking out smaller scale deployments and how do you see competition kind of sifting out here? Thank you so much.
Yeah, Bill. Yeah, great question. I think it would be crazy if we don't start to see other folks getting into that space, because the demand is obviously there. We didn't invent the demand. The demand is there, and now they're seeing it's easier to get to the 5 to 20 megawatt, not just for inference, but actually even for enterprise, even for regular compute. We're deploying, obviously, we're deploying five megawatt of B300s right now, right? So, or seven megawatts. So, the need is there, and they can use it. The question everybody's asking is, That's going to be affected like the big ones are, and I can tell you right now from being boots on the ground, I go to every one of those communities before we get engaged, and the 20 megawatt doesn't really move the needle for them. Now, they're interested. They want to understand it, but it's a lot easier to deploy in the 5 to 20 megawatt, and I don't mean easier, but it's easier to get past the political piece, and it's easier to find the 5 to 20 megawatt. You see a lot of bit mining sites that were out there, and people are going crazy over those. There's 5 and 10 megawatt available. The problem with a lot of those sites is they're not in heavy industrial usually, and you can't put generators. So there's tons of land out there right now with 5 to 10 megawatt that people could buy tomorrow. But if you do your due diligence, you find out you can't put generators there. If I can't put generators there, it does me no good. So they're out there. You've got to find them, and you want to make sure that they're zoned correctly and the community still knows what you're doing. They don't like it when you backdoor them at all, and they'll shut you down. We've seen that with another person that's deploying. So that's why our key is to get out there first and see if we're even welcome before we get engaged.
Operator
And this concludes our question and answer session. If your question was not addressed on the call, you can reach out to the company's investor relations team at duot at duotech.com. I would now like to turn the call back over to Mr. Recker for his closing remarks.
Thanks, Operator. Thank you all for the support in joining us today. We look forward to talking to you in the future. Have a great night. Thank you so much.
Operator
Before we conclude today's call, I would like to provide Duo's safe harbor statement that includes important cautions regarding forelicking statements made during this call. This earnings call contains forelicking statements within the meaning of the Private Securities Litigation Reform Act of 1995. Forward-looking terminologies such as believes, expects, may, will, should, anticipates, plans, and their opposites or similar expressions are intended to identify forward-looking statements. We caution you that these statements are not guarantees of future performance or events and are subject to a number of uncertainties, risks, and other influences, many of which are beyond our control, which may influence the accuracy of the statements and the projections upon which the statements are based and could cause DUO's Technologies Group, Inc.'s actual results to differ materially from those anticipated by the forwarding statements. These risks and uncertainties include, but are not limited to, those describing item 1A in Duo's annual report on Form 10-K, which is expressly incorporated herein by reference and other factors, as may periodically be described in Duo's filings with the SEC. During this call, Duo's also referred to adjusted EBITDA, which is a non-GAAP measure. A reconciliation to the most comparable GAAP measure is in today's press release, available in the Investor Relations section of the company's website. Thank you for joining us today for Duo's Technology Group's third quarter 2026 earnings call. You may now disconnect.