Prepared remarks
Thank you for standing by. This is the conference operator. Welcome to the Gorilla Technology Group, Inc. First Quarter 2026 Financial Results. Operator instructions: The conference is being recorded. Before we begin, we will read the forward-looking statement. Today's call includes forward-looking statements made pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These statements reflect management's current expectations and projections about future events and are subject to known and unknown risks and uncertainties that could cause actual results to differ materially. Forward-looking statements often include terms such as expects, believes, plans, anticipates, may, should and similar expressions. For a discussion of important factors that could affect Gorilla's results, please refer to our filings with the SEC, including our most recent annual report on Form 20-F. Except as required by law, Gorilla undertakes no obligation to update or revise any forward-looking statements made on this call, whether as a result of new information, future events or otherwise. I would now like to turn the conference over to Jay Chandan, Chairman and Chief Executive Officer; and Bruce Bower, Chief Financial Officer. Please go ahead.
Thank you very much. Thanks, everyone, and thanks for joining. Bruce and I will be direct today. Q1 was not a quiet quarter and it was not a neatly wrapped accounting quarter. It would have been easy to read a script, smile politely, slap AI onto the messaging and call it a day. I’m not reading from a piece of paper today. For me, Q1 was the quarter where Gorilla moved from turnaround into scale, and scale is rarely pretty in the early innings. Anyone who has built something meaningful knows you do not build a data center campus, secure power, buy hardware, deploy GPUs, hire people, expand products and move into sovereign AI infrastructure without creating noise in the P&L. If anyone expected a perfectly polished quarter while we build the next version of this company, they might also expect British sunshine to arrive on schedule — charming, but rarely accurate. Let me start with the facts.
We delivered $28.2 million of revenue, up 55% year-over-year. More importantly, we turned operating cash flow positive. Net cash from operating activities was $6.6 million versus cash used in operations of about $10.7 million in Q1 of last year — a positive swing of roughly $17.3 million, or 162%. We finished the quarter with a little over $98.4 million of cash, up 373% year-over-year. Put simply: revenue grew, our customers paid us, operating cash flow turned positive, and cash remained strong. This is real execution showing up on the cash flow statement, not market theater. The reported operating loss of about $41.1 million is misleading if you stop there. That loss was heavily distorted by two major items: a $20.9 million stock compensation charge that had been building for about 3.5 years, and $18.9 million of foreign exchange losses. Combined, those items represent roughly 97% of the reported operating loss.
Excluding them, the underlying operating loss was only $1.2 million. That’s the real context. This was an accounting-heavy quarter inside a company that grew revenue 55%, turned operating cash flow positive and ended with nearly $100 million of cash. That’s why I say this quarter separates accounting noise from operating reality. The stock-based compensation charge is noncash and reflects long-overdue equity compensation we’ve discussed with the market. Frankly, I would rather recognize the charge when our equity value is materially higher than to issue equity at depressed levels and dilute shareholders. Put simply, I’d rather take the accounting medicine at, say, $15 and not hand out the company at $3. This isn’t arrogance; it’s arithmetic. The FX loss was painful — nobody enjoys currency devaluation — but despite that, projects progressed and guarantees have been reduced. That is the operating story.
Now let’s talk about what Gorilla is becoming. Previously, analysts viewed Gorilla primarily as a Security Intelligence, Network Intelligence and Smart City technology company. That business remains core to who we are after 25 years, but we are moving into a much larger arena: AI infrastructure, GPU infrastructure, data centers, sovereign compute and secure national digital platforms. That transition costs money before it produces full returns. We are hiring people, buying hardware, securing land, advancing power planning, and taking colocation capacity. You may have seen our recent press release. We’re investing in GPUs, networking, storage, cabling, security infrastructure and operational systems. We could have managed the quarter for optics; we chose to manage the business for scale. The easy short-term move would have been to protect EPS, but our priority is to build the company. PowerPoints don’t run GPUs, headlines don’t cool data halls, and hope doesn’t secure power — execution does.
In India, we signed contracts with Yotta and materially expanded our AI infrastructure collaboration, creating a credible foundation for significant revenue scale. When I talk about Gorilla becoming a $500 million revenue business next year, I’m not guessing after a long lunch. I’m looking at signed demand, contracted opportunities and the infrastructure required to deliver them. Call it ambition with calculation. In Thailand, we are advancing a 200-megawatt AI data center campus in Korat. We’ve secured strategic land and the foundation for power planning; we are building the physical platform for Gorilla’s AI infrastructure. This is an owned AI infrastructure strategy in Asia — not just a concept, but land, power, planning, water, dark fiber, cooling, security and a real development path. Anyone can claim to build a data center; very few can assemble the infrastructure required to power it.
We’re also pursuing additional opportunities across Thailand, including Rayong. In Indonesia, we have moved forward securing colocation facilities in Jakarta and Batam. Across Southeast Asia, our goal is to combine owned data centers, colocation facilities, GPU deployments and sovereign AI demand into a regional infrastructure platform. I believe Gorilla has a credible path toward over 500 megawatts of AI infrastructure capacity by the end of 2028, and if we execute well, we can go further. Demand today is well north of a couple of gigawatts, so we must execute across Korat, Rayong, Bangkok, Jakarta, Batam, Singapore, Malaysia, the Philippines and other regional opportunities. Half a gigawatt of potential AI infrastructure is not normal for a company of our current size, and I’ve heard concerns that we’re too small to build it. That’s precisely why this opportunity is so significant for us.
Importantly, we are not becoming a one-dimensional data center company and we haven’t stopped product development. Raj, our Group CTO, continues to develop platforms, deepen our Security Intelligence capabilities, expand our Network Intelligence portfolio and push our sovereign technology roadmap forward. In Taiwan, we continue pursuing new customer opportunities. With Chelpis, we are advancing quantum safety and security, and with Astrikos in India we are strengthening our intelligence layer to predict and optimize infrastructure across cooling, IT load and physical systems. The future of AI infrastructure won’t be judged by GPU counts alone; it will be judged on whether infrastructure is secure, resilient, sovereign, efficient and trusted. Compute without control is just expensive heat. Gorilla’s advantage is building the infrastructure layer and the intelligence layer together. We are also investing heavily in people.
Over the last several months we’ve added more than 100 employees and over 200 contractors across delivery, engineering, finance, compliance, operations, commercial functions, procurement and more. That is not overhead for its own sake; that is execution muscle. Gorilla cannot deliver multibillion-dollar scale with a village-hall committee and a lucky spreadsheet. We are building the organization required for the next phase. So when you look at Q1, don’t view it as a small quarterly miss against an old model. See it as the first visible quarter of a company that is being built to be much larger. The old Gorilla proved we could turn around; the new Gorilla is proving we can scale. We are raising our full-year 2026 guidance to $160 million to $200 million, and I am focused on what it takes to build a profitable $500 million revenue business next year. That will require execution, discipline, capital, delivery and sustained effort across the Middle East, Asia and other strategic locations.
The direction is clear: revenue is growing, our customers are paying, operating cash flow is positive, and cash is strong. We are securing land and capacity, buying hardware, building data centers, developing new products, investing in people and building the capital platform to fund larger projects. This is not hype or spin — this is execution. In an AI market full of companies selling dreams, execution is refreshing. My message to the market is simple: Gorilla is no longer proving it survived; Gorilla is proving it can build something much larger. The market can debate the narrative, and that’s fine, but the cash flow statement has already begun to speak. Thank you. I’ll hand this over to Bruce, who will walk you through the numbers. Bruce?
Thank you for that. I think Jay covered all of the highlights. I just wanted to zero in on a few of those highlights and then a few other numbers that stood out to me. So the first is, as Jay mentioned, revenue up 55% year-on-year. You can see from the full year guidance, $160 million to $200 million is the range compared to last year. So that shows we're already on track with our year-over-year forecast. The other thing is that that revenue is converting into operating cash flow. We collected invoices from three large customers in the first quarter, which meant that overall net cash was $6.6 million. Subsequent to this quarter, we also got a release of all of the guarantees for our major project in Egypt. So basically the free cash portion of the balance sheet is very strong and the restricted cash, which a year ago was a very large number, has come down to almost zero. At the end of the quarter we were $98.4 million of cash and cash equivalents.
That shows, I think, that we have a fortress-like balance sheet which is able to tackle the projects that we have enumerated. So between the coax expansion into the colocation facilities and the projects at Yotta, et cetera, this is what gets us through the first stages. We also see the debt position continuing to perform in the sense that it is continuing to dwindle. We have $13.2 million of debt, which leaves us with a very strong net cash position. The last thing I would say is when you look at the top line and the operating cash flow, obviously the results are very exciting. We have invested, but a lot of this is operating leverage in the sense that the operating expense line shows up in the financial results as other operating expenses. That's basically the SG&A bill and it was only up 16% year-on-year. That's because some of the major hires we made last year and some of the major steps up in the budget we made last year are now paying off.
The second round of investments that we're making now into building out the infrastructure offering will soon pay off in a similar fashion. The last thing I want to talk about was really related to the FX losses and the stock-based compensation that Jay mentioned. There was a $1.1 million operating loss without those two big revaluations. I would note that we carry large balances in three currencies apart from U.S. dollars — in Taiwan dollars, in Thai baht and in Egyptian pounds — and given geopolitical events in the first quarter, all of those had adverse movements. Taiwan, Egypt and Thailand have all stabilized with their currencies, so we shouldn't see a repeat of that magnitude. Second, some of those exchange rate losses actually showed up in the operating figures because they had to do with the movement in the receivables value, which I think masks the underlying profitability of the business.
In a stable exchange rate environment, without significant geopolitical upheaval, we should revert to a much more positive net income profile. Many people have asked us over the last couple of months about our projects and the project financing update. Without going into too much detail, which I think lenders would not like me to do, we are very happy with the progress. We have multiple term sheets that we have either received and are waiting to sign and go into the documentation phase, or we are already in the documentation phase. The next announcement about the project-level financing will be when we say it's closed, and that will be the announcement that the project-level financing is closed for the various projects that will be funded. That is my update on the project financing; we're very happy with how it's progressing and it is comparing well with the assumptions we had when we went in and signed the projects. The profitability is there. That's all for me. I'll turn it back over to Jay, and we can open up for questions.
Thank you, Bruce. Operator, we're happy to take the questions.
Questions and answers
Operator instructions: Our first question comes from the line of Brian Kinstlinger with Alliance Global Partners.
This is Kevin for Brian. Could you talk about the planned timeline for the variety of HPC AI deals you've announced, including the multiple phases, the 3-year programs and the 200-megawatt campus in Thailand, as well as any others that I might be missing, particularly when each phase is expected to begin revenue generation?
Kevin, it's good to hear from you. Thank you. We have started our campus build-out in Korat. We have already started talking to the EPCs and are looking at water and power, so our build-out should start around the third to fourth quarter this year, which is when we will potentially start pouring concrete. Regarding other projects, Yotta has kicked off. We have placed orders with our OEM partner Supermicro through our distributor in India. We're working through customs and government of India regulations for import, which is a tedious task. That has kicked off already, and we expect our first delivery at the end of July. We have confirmation from Supermicro of the first delivery schedule. The second Yotta phase, which is the much larger project, is expected to be delivered at the end of August. After that, deliveries will continue each month through November. So revenues will hit our books with the first phase from September, and the second phase from October, November, and December.
In Asia, as you asked, we've just signed the colocation facility with NeutraDC over the last couple of weeks. That revenue is expected to hit our books from mid third quarter or the fourth quarter of this year because the data center and power will be fully connected. We have confirmed the full design architecture with the customer and are working with our partner Supermicro to get the delivery schedule. As of now the delivery schedule looks like August to September. We will keep the market updated as our timelines evolve. I hope that answers your question.
The next question comes from the line of Mike Latimore with Northland Capital Markets.
Congrats on the first quarter results. Your cash flow looks great. I guess you raised the lower end of your guidance from the start of the year, it was $137 million to $160 million. Maybe what was the main factor behind that?
Bruce, do you want to take that?
Yes, sure. So as you know, how we forecast guidance is we take what is contracted. We don't just stick our finger in the wind and think about what the pipeline looks like and hope for the best. So we feel confident in two things. The first is that the timeline, as Jay just mentioned, is looking very good for us to deliver above what was the previous low end of the range, $137 million. The second is that the second quarter and the third quarter are shaping up with more contracted revenue than we were originally planning on. So in between, by the end of the third quarter, I think we'll come out in a better place than we originally assumed. Those two factors led us to think the bottom end of this range needs to move up. Then the $200 million is still being ultra conservative. You heard, for instance, that one of the phases would be an October–November delivery. If there are any hiccups and it falls into the next year, I don't want to include that in our guidance for this year and then have egg on my face. It's much better to be conservative to the market, underpromise but be transparent. As things become 99% certain, we will adjust the guidance as appropriate. Jay, anything I missed?
No, I think you hit the nail on the head. Mike, good to hear from you again. Again, we are working very, very closely. As you can imagine, a lot of the global political environment in terms of deliveries and all that have also been a bit of a challenge. We are making sure that we're getting the right attention at the highest level at NVIDIA, making sure that we get it over with through Charles, who's at Supermicro, and make sure that we are able to get all the deliveries sent over to us. Now the good thing about India is that we've already gotten the delivery schedules, and that's why we upped the lower end of the guidance. Once we get through the hurdles over the next few days or weeks, we will come back to you with a more concrete, maybe an upgrade for the upper end of the numbers.
At that $200 million level, how much of that would be in the kind of AI, data center, digital infrastructure category?
Roughly around 60% to 70%. Our core business will continue to grow, but this AI is new, so we're going from 0% to almost 150% of that in terms of revenue. So yes, that's going to be where we are.
Then on the Egypt deal, you're at full implementation. Is there a recurring revenue that continues now?
Yes. So we have a 5-year recurring revenue, as we mentioned to the market about 3 years ago, post the completion. So we are looking to complete sometime mid to third quarter of next year. We're in the final implementation stage. As I've mentioned earlier, we've gone through the motions. We've done all the deliveries. Customers have been super happy. One thing I want to mention here is that we now have nil, near nil advanced payment guarantees on any projects. All our projects have been delivered successfully. The total advanced payments, I mean, as you know, 3 years ago, our advanced payments were well north of $50 million, $60 million being held hostage by our customers, which is obviously very important for them. So we can prove we're delivering. Today, it's $45,000. I just want to make that statement very clear. So it means we have delivered, customers have paid us.
That's great. And then just last for me on the gross margin. How should we think about gross margin for the year?
Yes. So last year's gross margins were in the low 30s. Given the growth in the AI-focused business, the margins will expand. The gross margins on the data center GPU-as-a-Service implementation are sort of 75% to 80% in a bad case and can be even higher. So that will drive the gross margins up for the full year. In this quarter, we saw a lower gross margin than we'd like given the mix were basically skewed a little bit towards more hardware. And then frankly, we are a little more aggressive on the pricing just to get an extra customer across the line. But overall, we've announced the contracts that will form the growth phase for quarter 2 to Q4, and the margins on those are much higher than the traditional business, the gross margins at least. So we expect them to move higher. When Jay alluded to a guidance update, when we update the guidance, we'll have a firmer picture with a pretty tight range on what that will be.
My apologies, and if I may add a quick point to Mike and Bruce: for Bruce and me this was a mobilization quarter. We've been front-loading costs. We hired a lot of people for a company of our size, prepared infrastructure, bought hardware, and run proofs of concept. We committed to data center capacity and had to pay significant amounts in advance. We also incurred project delivery and technical deployment costs. These expenses largely hit in Q1. Importantly, we're building capacity before the full revenue curve arrives. The concept comes first and gross margin recovery follows as utilization increases. You know data centers well, and AI infrastructure does not scale for free. We're ensuring the platform is up, stable, and meeting our 99.999% SLAs, and that our GPU deployments, data center revenue, and managed revenues materially improve over time. I just wanted to add that to Bruce's point.
The next question comes from the line of Bharath Nagaraj with Cantor Fitzgerald.
I think you mentioned 100 new people were hired and 200 new contractors. I guess that will only be partly reflected or maybe I think Bruce was mentioning maybe fully reflected in Q1 and you're continuing to hire more. So just wondering how much should we expect operating expenses to increase by in the coming quarters? And basically, that ties into any comments on where the EBITDA target should be for the coming quarters in the year? That's the first question.
Bruce, if you want to take the first half, I'll take the second half for that.
Sure. We mentioned that they were hired as contractors, which is one reason gross margin is depressed since many contractors are recorded as project-level costs rather than SG&A. SG&A was a little over $7 million in the first quarter and will increase in subsequent quarters as we continue to scale operationally, but it will not expand as quickly as revenue. Also, because we are adding higher gross margin business, gross margin should expand and flow through to EBITDA margin. Last year we finished with $101 million of sales and about $19.5 to $20 million of adjusted EBITDA. I expect margins to expand beyond that to 25% or 30% plus. We will provide the exact numbers when we have the final results.
Yes. And just to add to that for the second half, I think the market also needs to understand that the number of people we're hiring is not enough. This will expand by another fivefold or maybe even tenfold, both on the full-time side and among contractors. Let me explain. Today we bid these contractors because they're sitting together, putting all this infrastructure in play and so on. But look at the execution side of it. We need people for delivery and program execution, which we'll be hiring. We'll focus on engineering and infrastructure build-out. We'll be doing data center operations. To build a single data hall, we would need roughly around 300-plus people on average: each shift might have about 60 people, that's 180 people, and including everything else you're looking at about 300 people per data center operation. You've got GPU deployment, technical enablement, product development, SOC, NOC and managed services.
Then you have finance, compliance, procurement and project controls, and export controls. We have to have a separate legal team for all the export controls with the U.S. government and NVIDIA, which we have to support. And then finally we have to have our commercial support and what I call customer success. Now again, Bharath, we are not collecting these employees like stamps. We're moving from a lean turnaround business into scale execution. So the mobilization in Q1 and Q2 should be understood as hiring directly into what I call backlog execution. We're not hiring and waiting for new projects to come. We've already signed these projects. We're looking at roughly $3.2 billion coming from the year projects and another $2 billion of signed contracts, so you're looking at about $5-plus billion of backlog execution. Second, the Korat data center build-out is going to be at least another 1,000 to 2,000 people.
The India GPU infrastructure is up and running. We have the team from India here today in Asia, in Southeast Asia with us, and we are building all of our infrastructure teams. Then you've got Southeast Asia colocation capacity. We have outsourced most of that work to our friends at NeutraDC. But then you've got security, network intelligence and so on. So we are building what I call 0.5 gigawatt of ambition, and fortunately that is going well in our favor today.
That's very helpful color. Just a quick couple of follow-ups. In terms of the capacity of data center capacity or AI capacity you want to be installing by the end of this year, I think you mentioned 60% to 70% at the upper end of your guidance is to come from that. But in terms of the capacity, what's it going to be? I think it was historically around 100 megawatts, maybe that was at the lower end. So I just wanted to clarify what that number is for 2026? Because I think 2028, you have mentioned 500 megawatts.
That's a really, really good question, Bharath. We are aiming for between 100 and 150 megawatts by the end of this year. By the end of 2027, my personal ambition is to complete the full 500 megawatts. We've already received inbound interest in a number of other land sites and approaches from governments in various parts of Asia. We've received inward requests about how we can build up scale to about 2 gigawatts as well. These are conversations we're having right now. But my personal ambition, as I said, is that by the end of 2027 I want to have at least 0.5 gigawatts of power capacity, with a view that I've signed another gigawatt of development capacity as well.
Super. Just one, sorry, actually I have a couple more, if that's all right. Just on the GPUs, you obviously have a lot of competing demands. How confident are you that all the GPUs for these projects can be delivered given the supply chain issues? You've won a lot of orders and the pipeline is significant, so I was wondering about that.
Yes. Well, listen, if I had a magic wand and could look into a crystal ball, I would love to tell you I can have all this delivered by the end of this year and significantly boost revenues next year, but it takes time. NVIDIA is releasing a lot right now; if you look at NVIDIA's release, the next generation of Vera Rubin is also coming out. Customers are keen to look at that and potentially talk to us about it, and that can change the entire goalpost sometimes. So we're making sure customers' architectures don't change and that customers stay grounded. There's a shiny object out there and customers want to run toward it, so we've got to keep them grounded. On the delivery side, fortunately, we have not had any major issues with NVIDIA to date. The global concerns today feel like a noose around my neck, but they haven't created a major problem yet. What has created some delay is the current lack of availability of memory and storage in the market, and now we are also seeing CPU shortages.
We are working with our partners and are very closely integrated with Supermicro right now, working day in and day out. I was there the whole week before with them, and we're spending the next full week at Computex in Taiwan, where we are sitting together and planning how to ensure memory supply. Getting the GPUs is great, but our capacity needs to increase, so we are working hand in glove with every major partner across the region to make sure it does not falter.
One small and minor accounting question. On the SBC costs, am I right in saying that given you recognized most of what you had said you would in Q1 itself, the remaining quarter should be minimal. Is that right? Or am I getting that wrong?
Bruce?
Yes, I think that's correct. There was some deferred stock-based compensation that has been outstanding for a couple of years. For various reasons, we decided to pay it this quarter, so it is no longer an overhang.
Bharath, if I may add to that, and this is not me being funny. I've seen comments like, oh, my God, CEO has gotten paid and blah, blah, blah. No, this is not just the CEO. This was for the employees as well and everybody else around the company. I want to make sure that the compensation, the market understands that the compensation was due for the last, what, nearly 4 years now since the company went public. And employees need to be paid. They need to be given their stock. Unfortunately, it had to come in this quarter. That's okay. If I don't pay my employees, that's the wrong thing for me to do. So I'm setting the right precedents.
The next question comes from the line of John Roy with Water Tower Research.
So Jay, obviously, there's been a lot of talk about much larger and larger projects, AI infrastructure, GPUs, data centers, et cetera. I was curious, I know Bruce talked a little bit about funding just to your maybe philosophy about how are you going to fund these massive projects? And where do you stand on that? Maybe just give a step back and tell us where you're at.
John, good to hear from you. I was wondering when you would ask me a question. That's a very fair question, and frankly, it is the right question. The scale of Gorilla has changed. You and I know we talk regularly. We're not talking about small software deployments. We're signing and pursuing large AI infrastructure, GPU data center projects across India, Thailand, Indonesia, Malaysia, Singapore, the Philippines and so on. That requires capital. And there's no version of my story or this story today where we signed multibillion-dollar opportunities to buy GPUs, reserve data center capacity, procure networking, memory, storage, secure power, buy land and build data centers without funding the business properly. GPUs take money. I don't know if people realize buying a B300 server costs me more than $0.5 million, excluding networking and all the other related costs. So when a customer tells me they need 1,000 servers, you're looking at about $500-plus million of investment just on the GPUs for the servers.
Then you've got networking and other items, which costs another $20 million to $25 million. It's a pretty penny. So where are we today? We have not yet relied on dilutive equity to fund the build-out. Our approach has been to protect shareholders while building the capital stack required for our platform. We are actively working on vendor financing. We have received term sheets in the range of approximately $0.5 billion to $1 billion across vendor financing and debt structures. We are progressing with various debt financings. We have term sheets and bank-led proposals between $300 million to more than $700 million, $800 million that contemplate lending at the project or SPV level rather than relying purely on the listed path. You remember what Bruce said last quarter: we're making sure it's nonrecourse. And I think people need to understand when Bruce meant that, he meant it for real. So we're also looking at different levels of SPV structures.
That is important because infrastructure assets can be financed against their own cash flows, contracts, equipment and project economics where possible. So we are working with various SPV-level structures. We're also building Gorilla Capital. The market seems to have forgotten about it because it's what I call a strategic funding platform. The goal is to bring long-duration capital, including pension funds and endowments, institutional investors with structures that can support a seven- to ten-year long-life infrastructure asset investment. We're matching funding to the asset. GPUs and data centers and contracted infrastructure revenue should not be financed with short-term thinking. The capital structure has to match the commercial life of the assets. More importantly, we're being very disciplined about shareholder impact. We will not do financing simply for the sake of financing. The objective is growth, profitability and shareholder value.
So what the market needs to understand — I have no idea what's happening to me, sorry, I apologize — is that we're looking at potentially more than $5 billion of signed contracts and executable opportunity across our AI infrastructure and data center pipeline. The market knows about this already. If we want to move Gorilla from a $100 million revenue business last year to $500 million revenue next year plus annualized business for the next five years, then the business has to be funded like a serious infrastructure platform. Growth requires capital, and profitable growth requires very disciplined capital. So that's why we're not raising money because the business is weak. We're assembling capital because the opportunity ahead of us is much, much larger. My message to you and to the entire market and to everyone listening is that we're funding growth through a variety of vendor financing, SPV-level financing and long-duration institutional capital. But we're doing it very carefully to protect our shareholders and keep them in mind at every step. I hope that answers your question.
Yes, it does, actually. It kind of brings up a corollary question, which is the pipeline. Can you give us any kind of color on the pipeline? I know there's some big numbers out there. Just curious if maybe you could summarize it with some...
Sure. So today, the pipeline, the signed contracts or what I would call the backlog, is well over $5 billion, okay? The pipeline to be signed or in negotiations and discussions is well north of another $5-plus billion. That's excluding any of the build-out we're doing currently in Korat or in Rayong and so on and so forth. When I mentioned this previously to Bharath, I made this very clear to him that we are looking and my personal ambition would be to get a full gigawatt in there. If I get the full gigawatt with offtakers, and by the way, just FYI, we do not sign any colo. We're not purchasing any land without an offtaker. I have signed offtakers completely ready to take over the capacity on day one. I will not spend on GPU power without having an offtaker for a single hour. So our revenue will hit the books as soon as the date opens when the ribbons have been cut. So if we do the one gigawatt, then you know what the revenues are. I'm not going to provide any numbers right now, but we will look at a significant upgrade from even the $500 million plus number.
The next question comes from the line of Barrett Boone with RedChip.
Jay, Bruce, congratulations on the strong start to 2026. As discussed earlier in the call, receivables came down meaningfully during the quarter. Can you talk about what's driving the better collections? And how we should think going forward about cash conversion as we scale towards the $500 million revenue target?
Sure. So what's driving it is really we have 3 core customers, which we disclosed in the 20-F. And we delivered over the course of 2025, we invoiced and then in 2026, we said these are terms, make sure that we collect. So 2 of them have always been extremely prompt payers, and that's the kind of customer we like. And then in the third one, it's really just a simple logic, commercial logic where we say, look, in Egypt, we've been working together since July 2023 when we were awarded the contract. We've come this far. We've done this much for you. Is it too much to ask that you pay on time and the customer is recognizing the value and then the core nature of the infrastructure that we've built. So that is helping and just the sticky nature of our product. And then the thing I would say is that with new customers, we're extremely vigilant about the payment terms for who we onboard.
Absolutely. Thanks, Barrett. Bruce, that was actually quite interesting. You stole everything from me already. So just to add to it, Barrett, personally, revenue is wonderful. Of course, everyone likes revenue. But cash is what separates the business from being a brochure. We produced operating cash flow. As you know the numbers, I'm not going to repeat them. There was a huge swing: a positive $6.6 million and about $1 million in other changes. But more importantly, what drove it? I think we need to educate the market. First, our customers paid us. That may sound obvious, but when you look at large infrastructure projects, payment behavior is one of the clearest signals of delivery quality. It shows the quality of the company. Customers do not release meaningful cash because they're feeling charitable; they release cash because milestones are being met, documentation is being accepted, and projects are moving forward.
The second part is that we've also tightened our project discipline. We're no longer a $20 million revenue company. We're being more aggressive internally on invoicing, collections, milestone tracking, project governance, and customer acceptance. It is not enough to win large programs; we must convert that into recognized revenue and subsequently into cash. The difficulty is that we want to make sure that we are running the business profitably. The third most important part of the business is that now we're building the business where cash conversion is becoming part of the operating model, not an afterthought. What matters next is scale. If we are serious about moving toward $500 million of revenue, we cannot allow working capital to become a museum of unpaid invoices. We need disciplined contracting, delivery, acceptance, billing, collections, cash application, and so on. As we scale Barrett into data centers and GPUs, cash flow will not always move in a perfectly straight line.
I wish it did, but these are all large programs. We're going to stick to our guns every single month. But the Q1 signal is very important. We grew revenues. We reduced our receivables. We reduced our advance payment guarantees. I just mentioned this earlier to Mani. It's gone down from $50-plus million to $45,000. That's the next phase of our business evolution. So the simple answer is, as we scale forward, cash conversion will become the key metric for us, which Bruce and I are watching and will continue to watch like a hawk. Revenue gets attention, but cash earns respect for me. So for me, cash will be our standing ovation going forward, Barrett.
Understood. I just had one last question. Actually, about today's release, you do cite that the combination of infrastructure and AI products gives Gorilla leverage. Can you talk about how everything sort of works together and how these products help you win infrastructure deals that perhaps a pure-play data center competitor couldn't?
That's actually a really good question. I think, again, markets and many, many, many investors have also missed this. The simple point is, Barrett, we're actually not just selling space, power and cooling. A pure-play data center operator will give you building, they'll give you racks, they'll give you power, service desk, useful, but it's not stuff of Shakespeare, okay? I'm just using a British chronology here. Why? Because Gorilla brings the full operating layer around the infrastructure. Think about it this way, site assessment, power planning, cooling infrastructure and architecture, feasibility studies, data center readiness, all that is being done by us. Racking, stacking, cabling, GPU commissioning, network integration, cloud enablement, that's also being done by us. When you look at security, whether it's physical security, access control, cybersecurity, your SOC capabilities, your NOC monitoring, CCTV, access controls, we build and manage everything ourselves.
We also operate it. That means we have a 24/7 monitoring, managed services, remote operations, preventative maintenance, life cycle support, all of that. So for us, we're not just providing a room with very, very lovely blinking light. We are making sure that our products are sitting in it. And that's why we invested into Astrikos, right? Look at the difference. We have security intelligence products, which actually help customers protect their critical infrastructure, their endpoints, their users, their cameras, their operational networks. We've got the network intelligence products coming in from our friends at Astrikos. We have built our own SD-WAN, secure tunneling, orchestration and edge connectivity for products. Our business intelligence layer talks about all of our operational data, our infrastructure data, our video, IoT analytics and actual decisions. So when we sit with a government, telecom operator or enterprise and any infrastructure partner per se, right, we're not saying here is a building, good luck.
That's not us. That's not a strategy. That's a real estate with electricity, okay? It puts Gorilla in a much stronger position than a pure data center competitors. So with Gorilla, it gives them capacity plus control. So look at it this way, customers care about sovereignty, what is it, security, latency, compliance and so on and so forth. They don't have 12 vendors doing this. Typically, when you go into a data center, you've got 10 to 12 vendors doing this. We are there for the one throat to choke when something goes wrong. This is very, very, very important. Now our model also gives us leverage. It gives us scale. It creates differentiation. It also creates what is called as the Gorilla edge, right, all pun intended. Because if you look at the full stack model across design, development, deployment and operations and so on and so forth, we're sitting right at the top of it. So if you look at a pure data center, think about it as someone giving you a garage.
But more importantly, Gorilla gives you the garage. It provides you the engine, it provides you the security system. It provides you the control room and someone who's awake at, let's say, 3:00 a.m., like I was awake at 2:00 a.m. this morning, when things actually matter, Barrett. I hope that answers your question.
That concludes the question-and-answer session. I would like to turn the conference back over to management for any closing remarks.
Thank you very much, Pamela. I really appreciate it. Analysts, investors and employees listening to my conversation today, thank you very much for your support. I would want to leave our investors with this thought. We have rebuilt the business. We've proved the technology. We've collected cash. We've turned operating cash flow positive and are now moving into a much larger arena, AI infrastructure. We've always been an AI infrastructure company, okay? We've been building the blocks. If you hear me what I said in my first interview on the NASDAQ in July of 2022, I said we were moving into building an AI infrastructure platform as a Service. That's exactly what we're doing. Data center is a part of it. It's not a pivot. So please do not use that word. We're not pivoting. We are building the platform, and we're going to close and secure the platform. We're building GPU capacity. We're building sovereign compute, and we're making sure that national platforms function.
We're buying land. We're securing power. We're taking data center capacity. We're building new products. We're hiring new people, more people needed to deliver. So we're not talking about scale from a distance, we're actually building it. So my message to the market is very simple. Judge us on execution, judge us on cash, judge us on delivery and judge us on whether we keep scaling. Everything else for me is commentary. And frankly, there's been plenty of commentary from people who are just sitting on the sidelines, and these people are not even able to build a sandwich, let alone an AI infrastructure business. So Gorilla is not only getting started. The market can doubt the story if it wants, but it cannot ignore our direction of travel. Thank you very much for listening in. Have a lovely evening.
Thank you. And this concludes today's conference call. You may now disconnect.