Good afternoon, everyone. Now presenting will be Danny Chism with TSS.
Appreciate it. Thanks for everybody coming. Run a quick run-through, and then as usual, open it up for Q&A at the end, assuming I don't get too verbose along the way. Quick disclaimer, obviously all the stuff you can read on the website, too, and all the investor relations stuff, everything's subject to future revision, all that good stuff. Core, what does TSS do? We, for many years, were in the business of designing, building, and deploying actual data centers. We figured out pretty quickly, we, I say collectively, because I wasn't with the company yet, but years ago decided that was a feast or famine business that we didn't really want to be in. We stopped building and designing data centers. We really migrated much more into integrating computer racks that go into data centers.
Some other ancillary activities that go on around that, but that's the core business. We already were having pretty good results with that. Built a great relationship with, we don't name them publicly, but one primary customer that we work with, an ODM that you would all absolutely recognize their name. Their name may be on many of your laptops. Anyway, do a lot of work with them. When AI was really starting to boom a couple of years ago, they already liked what we were doing and asked us, "Hey, would you like to do this for us for AI computers as well?" Absolutely. It's about the time you see our stock really take off. It's been a really nice tailwind. We really started to take off a year or so before that. Had a new CEO, new COO come in about three and a half years ago.
Really turned the company around, rebuilt a very strong relationship with our primary customer. They don't necessarily call us a partner because they're pretty cagey about that name, but it really is a great partnership on both sides, trying to find ways for both of us to make more money. Around that, computer rack integration, if you think of it, I think of that as somebody else makes all the LEGOs. They send us the LEGOs. They ask us to put together the LEGO set. We put it all together. We plug it in. We run it. We try to break it for six of 24 hours. If we can't break it, all the GPUs work, all the fiber optics work, we load it on a truck, and it goes to a data center somewhere.
A lot of the hyperscalers or other data center operators that you would absolutely recognize their name, high likelihood their racks came through our factory. Many of those end customers don't even know who we are, because our customer, again, is the ODM that's selling the completed rack to them. Around that, we also do strategic procurement. For that same customer, and I'll pause there, this customer represents 99% of our revenues. Yes, we recognize that's a high concentration overhang on the stock. Some part of the strategy is to diversify away from not shun that business, continue to grow it, but try to grow the rest of the business with other customers as well. A few strategies we'll talk about later about that.
Procurement business for that same customer is typically buying their servers, not AI servers, just one server that may be going somewhere or 10 servers that may be going somewhere. They may need us to load a software image on it or put it in a ruggedized box or do something else to it to transform it somehow, maybe a different memory card in it that that particular customer wants something different than what they typically manufacture. They'll send that to us, have us do a little bit of tweak to it, and then send it out to their end customer. If we do that, we report the gross value of that transaction.
Instead it's just a buy and sell, if they just need us to go acquire some software from somebody else and sell it to their end customer and they don't have that relationship, we'll just earn an agency fee on that, and that's what we record as our revenue. That's the procurement business. Then the third, really deployment and configuration and modular data centers go together. We call that our facilities management business. Largely modular data centers. Think about a 40-foot shipping container, six to eight racks of computers in there. It's a little self-contained data center. All the battery backup, air filtration, everything that you need, even direct liquid cooling if you need that. We do that in modular data centers. We've deployed over 500 of those.
We continue to maintain a fair number of those today and earn, not warranty, but maintenance revenues off of that. The other couple of things I mentioned on here, just one that really our primary customer really values our flexibility. It's one of the keys that sets us apart. If they've got a solution that they need to provide to a customer and they can't figure out how to do it, many times we can be much more flexible than they can and go make it happen for them. Investment thesis is really, it's a big market. It's growing. There's huge tailwinds. Probably need to update or comment a little bit about being in the early stages or earliest stages. AI's in full swing. I think it's still got a really long tail on it. I don't think that'll be a surprise to anybody in this room.
Not necessarily in the earliest stages. Marquee end customer that I talked about, high demand offering, multiple growth levers. Some of that we're looking at, again, how do we diversify our revenue stream? Some of that may be new offerings. Where our earning process ends today when that rack ships off to the data center, how do we follow that rack into the data center? Are there services we can provide for inter-rack cabling, services in the data center? Those are all some of the things that we're looking at. How do we expand into that? We don't do that today, not in any meaningful manner. That's one that last August we raised a little bit of capital, about $55 million, largely to be able to seize those opportunities.
We could build that ourselves and grow it over time, but that might be four or five years to make that happen. More likely, and probably a better path to accretion for our shareholders, is to go acquire somebody. We've been looking at a number of companies. Nothing to announce today on that, but something that we continue to be active in those discussions. Expanded capacity, where we do this today primarily with one customer. We could go potentially put up another factory for somebody else, or a factory that could serve three or four other customers, depending on how we structure that deal. That's a big opportunity for growth. We don't provide revenue or EPS guidance. We have guided for EBITDA. We did about, I think, $18.6 million last year.
We've guided $20 million-$22 million this year, and last quarter, we guided to the upper end of that, looking more around the $22 million range. It's big. That's what this whole slide says, right? The TAM is huge. We get a small piece of that, but it is immense spending. Again, no surprise to anybody in this room with all the AI data spending. The three lines of business. Procurement, these are largest revenue, smallest margin. Two on the right side of this slide with the facilities management being the lowest revenue and the highest margin. These businesses tend to run procurement, a blended average of around 6%-7% margin. systems integration, historically, had been about 25%. That's more around a 35% margin. I think last quarter was 37.2%, but I'd expect that around 35%.
facilities management tends to run about a 50% margin business. Here you can see the impact of that. In 2025, you can see facilities management was only 3% of the revenues, but it's about 20% of pre-tax income. Same thing with the systems integration, 17% of total revenues, but 22% of the pre-tax income contribution. While procurement is a meaningful part of the business, we think of the facilities management and the systems integration as the meat and potatoes of the business. Procurement is really nice. It's tasty gravy to go on top of the meat and potatoes. It's not the critical business. It's one we absolutely love. It's a lot of revenue, helps grab some headlines, but that's not really what's going to drive the huge valuation of the company moving forward.
The biggest opportunity there is in the rack integration and some of the facilities management. This will give you a quick view, again, a different view of the smaller kind of $1.3 million-$1.4 million run rate in facilities management. We were running about $9 million per quarter in systems integration. We signed a new agreement in Q4 2025 with our primary customer, again, for the systems integration on AI. That took into account the fact that we put in additional CapEx and took on more power at the building, which they needed us to do in order to be able to continue to integrate the next iteration of racks, progressing technology, more demands on cooling.
We spent more CapEx on that based on their needs, but that wasn't yet reflected in our underlying agreement, which is structured to help us recapture all those costs over a number of years. That was addressed in Q4. That's where you see the revenue spike to about $14 million in Q4, a little over $14 million in Q1. This last quarter, we announced that we will be investing again, because of their needs, another $17 million or so between here and Q3. I'd anticipate that systems integration revenue to jump a little bit again as that starts getting reflected in what we're charging to our customer once we put that into use in Q3. In procurement, again, 6%-7% margin business typically, although huge volume. We typically run about $30 million-$40 million per quarter.
If you look at Q1 last year, that's the one that the banner even cuts it off halfway up. It was about $90 million Q1 of last year. When we reported this quarter at $40 million, that was a good, strong quarter for us for procurement, but the comp against $90 million was ridiculous. The Q1 last year, the headline was total revenues down 44%, stock tanked a bit about after that. The fundamental investors, the institutional investors that I've talked to, were saying, "It was great. We loved it. It was a good buying opportunity, dollar cost average and lower." You've seen it lift back again. We went down, I think, kind of in the $11 range or so, back up a little over $13 now. That was the biggest challenge in Q1, just in the reporting.
When you look at the fundamentals of the business, everything was hitting. Strategy. Again, how do we diversify the revenue? Some of that is doing more with our existing customer. Some of that is I'll back up. With our primary customer, we really can't go do the rack integration for direct competitors. But if we were to have an opportunity to work directly with a hyperscaler. If you think about some of the big, the Metas or Googles or AWSs of the world, typically they're not buying from the ODMs, the HPE, Dell, Supermicro. They're typically not buying direct with them. They're buying from the OEM. They're having those servers built directly for them, that's what's called white box. There may be some opportunity there, right? That's one potential avenue for expansion that doesn't take food out of our customer's mouth.
There's no issue there that we would be competing with them. Intense customer focus. We continue to cultivate the relationship we've got, continue to try to grow it, make it better. At the same time, how do we go get other customers, and how do we follow that into the data center? Those are some of the key growth areas. Again, we raised the capital last year specifically for growth, whether that's M&A activity, whether it's more investments in CapEx to continue to support this customer and their growing needs. The structure of our agreement is a little bit unique. It does have a multi-year term, and we've structured it such that we are comfortable that we will fully recover not only the CapEx, but then some profit on top of that CapEx and make sure we're covered from our cost of capital.
The same for if we've got a few hundred people sitting around waiting to work on stuff, we've got some minimum guarantees there that if we're sitting there with nothing to work on, we still are at least made whole plus a small profit. This slide, basically, it's up and to the right. That's the whole theme of it. Company continued to grow. Q 2025, I would say, was inflated a little bit by that first quarter of the $90 million of procurement. Again, typically that runs $30 million-$40 million. I'd love another $90 million quarter, but that's going to be the rarity, right? There will be some that are $20 million. There will be some that are $50 million or $60 million. The $90 million was really an outlier. The 245 also is going to be a tough comp for the entire year.
That's inflated by the lowest margin business. I'm not concerned from an earnings standpoint. Operating income, you can see down a little bit in Q1, $2.3 million versus $2.6 million. Again, impact of the same large procurement activity in Q1. Everything else as we looked, I'll flip back a few slides. Back on that one, you can see procurement still in that upper end of the $30 million-$40 million per quarter range. Systems integration's still really high. Facilities management, still good, strong, about $1.3 million. The anomaly there is Q4. We had some discrete projects there where somebody wanted us to refurbish some of their modular data centers that were out there. That's the spike to $3.5 million saw in Q4. Diluted EPS, same impact of that Q1 comp. Adjusted EBITDA, even with that tough comp, we still pushed adjusted EBITDA up from $5.2 million - $5.3 million.
As we get more reasonable comps in Q2 through Q4, I expect that comparable year-over-year to look even better. The other thing that you'll see if you dig into the details, at Q4 of 2025, we completely reversed, or almost completely reversed the valuation allowance on our deferred tax asset. Basically doubled our net income. Instead of roughly $7.5 million, we reported about $15 million of net income. Which is great because it juices that up, because we had that valuation allowance before, our effective tax rate that we were reporting was like a 1.8%, right? It was all getting absorbed by that DTA. Now, even though it's no difference in cash taxes that we're paying, we're recognizing the full tax provision. That hits the EPS line, even though the true economics are unchanged.
Recent business developments, as I mentioned, we signed a long-term agreement with a primary customer. That was in September of 2024. Fast-forward all the way over to the right, we just recently enhanced that, took into account the additional CapEx, took into account the additional power we're bringing into the building. The other thing we did with that was we extended it an additional two years on top of the multiple years that it already still had to run. Good runway for future growth. I know a lot of people in the AI space specifically are concerned about, "Hey, is this a bubble? Does it go away quickly? How long does it last?" We've got really good comfort for a number of years here. I personally think we'll hit another refresh cycle before too long.
Otherwise, the people who are adding all that compute today are going to get obsoleted. I think it doesn't go away anytime soon. November 2024, we uplisted on Nasdaq. We were OTC prior. May 2025, we relocated our current facility. We were in only about 100,000 sq ft in Round Rock, Texas. We had 2.7 MW coming into the building. Great for what we did then, but knowing we were going to start doing AI racks that were more power hungry, also takes a lot more power to cool them, we knew we'd be obsoleted if we didn't move. We moved about 9 mi up the road in Georgetown, Texas, 213,000 sq ft, a little bit more than doubled our square footage. More importantly, we went from 2.7 MW - 15 MW of power in our building today.
The city's already committed to giving us up to 40 MW if we need it. Don't need that today, but that is ultimately going to be a limiting factor for some people in the AI space, and that's one we've already solved for. A couple of new additions to our executive team. Again, where we talked about, hey, M&A being a potential opportunity. Matt Wallace is not only M&A, but he's got a lot of M&A experience. Came with some good experience within Dell, so he's familiar with the market, and has worked in their strategy area. Chief Strategy Officer, a lot of M&A focus. David Hull also spent 29 years at Dell. Actually designed a lot of their original AI racks. Designed a lot of the modular data centers when they were first getting into that business.
Tremendous amount of knowledge, tremendous contacts within a pretty attractive potential customer with Dell. I won't hold it against him they went to Texas Tech, but, you know. Anyway, both really sharp guys, great additions. These are not replacements. These are newly created positions. Specifically knowing if we're going to do some M&A activity, let's have the right people in. Let's make sure we're not just talking about diversifying the revenue. Let's go make sure we can execute on that. David also, as we talk about following the rack into the data center, or diversifying the number of people with whom we're doing the rack integration, David can be critical to that. His skill set and ability to recruit in talent is going to be tremendous. Some of the recognition for what we were doing.
One of our customers, Dell Technologies, awarded us the 2024 Best Deployment Partner. Big honor. Tons of people competing for that, and we won that one. Got to beat our chest a little bit. Don't give me too much grief. I liked the hair when it was still brown and fewer wrinkles, so yes, my picture's about 20 years old, but I refuse to give it up. The other guys actually look like they do. That's the rest of our key executives other than, we also have Janet Morrison, who's our Chief People Officer. Tremendous asset. When we first started doing the AI rack integration, our chief operating officer and Janet hired about 150 people in six weeks. We were an 82-person organization at that point. We can be nimble, we can execute really quickly. Quick view of the balance sheet.
I just got the sign that we've only got five minutes left, so I want to leave time for Q&A. Basically, we're less than a one leverage ratio, pretty strong cash position. We've got the ability to execute. There is about $4 million of current debt in the current liabilities line. The long-term debt you see at $13 million is really more like $18 million or $19 million. When you look at the cash going from $85 million- $65 million over the last quarter, pair that with the total current liabilities going from $73 million - $41 million. We had a lot of procurement activity going on at year-end, and we paid for it in Q1. The cash was used to pay down the AP. We've already gone through that one, I won't hit it again. That's kind of it.
Would love to open it up for Q&A if anybody's got questions. The question was whether we're seeing much demand in the modular edge part of the business. Not as much as we'd like. Again, it's a 50% margin business. I would love to see more of that. We've seen some interest in it, we don't typically sell those directly, right? We do it with our partner, they typically sell it, and then we actually do the work behind it. There's been more interest. The challenge there has been largely waiting for the container itself. It may be a year between the time that you order it and you take delivery. If you're putting AI in that, you may be two chipsets behind in technology by the time you take delivery. That's been the biggest challenge.
If they can ever solve that, where you can get it within four or five or six months, I think we'll see a pretty good tick up in that. The other one I expect is, we'll probably have a little more interest from enterprise with the modular data centers because they can basically have a fully functional data center with six or eight racks. Is decreasing the customer concentration a focus for you guys, and if so, what is the timeline for doing so? Absolutely, that's a focus. Again, we want to lean into our relationship with our primary customer. We don't want to shrink it. We absolutely want to expand that beyond that customer. That's been some of the focus of the M&A conversations, some of the stuff about how do we get into the data center?
Can we do rack integration for somebody other than that primary customer? Timeline's a tougher answer. Some of the reason we raised the capital when we did was we've had some active conversations with potential targets. Generally, what we've seen is expectations on valuation have been beyond what we are willing to go to. You don't want to see a company raise money and then not put it to work. Worse than that would be putting it to work on something that we're not disciplined and chase the deal. That's when we maintained that discipline and said, "No, we've got to stick to the correct multiples." Great question. The question was, do we actually have the ability to extend to other customers? We got plenty of power.
Right now we could do probably three times the volume in our current facility as what we do today. From a competitive standpoint, the structure we've got with our current customer, they're basically making us whole for that whole building and all that ability. We probably could not, from a competitive standpoint, do that in that building. If we got another large one, we'd probably need to go put up a second facility, which would be great. We'd make sure we structure a deal similarly to where we make sure we've got the security that we're going to recoup all that cost and cover the rent. Yeah. Yeah, good question. How complex is rack integration, and could our customer not just bring it in-house? They actually do it in-house. We are one of two facilities they use for systems integration, us and their own.
Their own is in Massachusetts. A lot of it, the logistics, we end up being a cheaper solution for them many times, depending on where the end customer is. We think we do a really good job of it as well. Is that time up or is that one minute? Time up? Okay. Hopefully that answered it, sort of. Yeah. I don't think it slows down the development. Frankly, we work very hand-in-hand with them on improving processes, right? We trade information with each other. We've been to each other's factories. It's a very cooperative relationship. Yeah. Cool. Thanks. I don't want to hold up the next presenter, appreciate everybody coming in.