Hardware and brings it to industrial spaces, almost exclusively on the commercial side. The majority of the discussion you'll hear us, hear me talk about, will be more centered and focused on strategies and growth in this OSS segment of the market where we're ruggedizing the enterprise-class capability and taking it to the edge. That's where our larger growth and our substantially higher gross margins are. This side of the business is, Bressner's generally a low 20% gross margin business. On this OSS side of the business, as we're growing, we're able to attain gross margins of 40% or greater. So if I can put us in context where you see us or where we operate, if we look at just high levels of compute performance with no ruggedization, at the top end of that is your data centers. That's where you find NVIDIA, Dell, Supermicro. We don't operate in that market.
A lot of players, very price competitive, large manufacturing volumes. If you bring the performance down and you make stuff so rugged a Marine or a soldier can deal with it and not break it, you may have heard of companies such as Curtiss-Wright or Mercury. They're doing very ruggedized compute out on the edge. We're in this unique top-right corner where we bring the fullest power, the highest levels of the performance of the data center, but we ruggedize them so that even a Marine or a soldier can operate them, drag them through the mud, kick them, run them over with tanks, and the hardware will still continue to operate. The company's built on a couple of things I'm going to show you here and some other charts, but this growing market for AI/ML, sensor fusion, and autonomy is really taking off, right?
You can see it's almost growing on an exponential level. Much like I've come to find, the internet is probably going to stick around. I think we can all agree that the likes of artificial intelligence, machine learning is likely to stick around for a couple of years, and you're going to see why that's important to our strategy. I mentioned the billion-dollar pipeline that we're working off of. As we factor down and look at that, we think we've got more than enough capability in just domestic United States operations to be able to grow the company organically at 25% a year. And then I'll talk into the idea really is to generate this pipeline into platforms. You become incumbent on those platforms in defense and commercial markets, and then they represent almost annuities as you continue to service that platform, tech refresh updates, sustainment, and support.
For microcap company , we're pretty strong on the balance sheet side. We have next to no debt, close to $12 million of cash sitting on the balance sheet. So operations of the company are fairly strong. We've just gone through kind of a new management overhaul and board reprofile. I joined the company about 18 months ago to bring an increased focus onto the defense market, to grow the company not only in commercial but also in defense. The company traditionally has been more 70%-80% commercial. We're working more towards a 50/50 split for commercial and defense as the company continues to grow across both markets.
The underlying theory or opinion behind all this strategy is, if I get your attention here to this graphic on the right, if we go back about 2017, 2018, commercial and defense markets were using open systems architectures to implement computing at the edge. The instantiation of that is through an architecture called embedded compute or a capability called VPX, so if you take—if you have an iPad, if you consider a computing card about that size with a high-end Intel CPU on it, that would be embedded or VPX. What wasn't envisioned out in 2017, 2018 when this started was that architecture can really only support running one or two AI inferences for a given card or chipset and for that volume.
As I mentioned, though, today, if we look at what's the demand for AI/ML even on existing platforms, as people are finding out with the data from the sensors, with the data from the operation, they want to run more than one or two AI/ML inferences or autonomy inferences inside of a single platform. So the demand is really to run 10, 12, 15, 20, 30 AI algorithms in a given platform or in a given volume for compute. As you can see, the current technology can't meet that. So what our company offers is this green line. We take the power of the data center and the enterprise, and we move it forward to the edge compute area in principally the same space and volume and principally the same price point as this yellow line.
Why that's important is this is one of those architectures in life that it's not a question of if people are going to move to the green line, they have to. You can't take advantage of anything that's autonomy, sensor processing, sensor fusion, or AI/ML at the edge with this blue demand increasing unless you start moving up into these enterprise-class architectures. This is kind of further supported by this bottom bullet down here. About 10 months ago, the Department of State is restricting GPU sales to the likes of China and Russia. What that means on a defense application is the Russians and the Chinese understand the power of this green line and have been implementing AI/ML technologies with that capability dating back a couple of years. The U.S. and our allies are stuck on this yellow line.
As you can see, there's an impetus for our defense forces to move up to the green line to catch up and get ahead of our adversaries. What that pipeline and what that chart mean for us, right, is why I mentioned why we have such a large pipeline. That billion dollars really breaks out for us into about $690 million in defense and $525 million in commercial. Those are predominantly all domestic U.S. opportunities for the company. Having run broad defense and commercial global portfolios in my past, I've come to find out that generally there is pipeline opportunity equal to what you have in the U.S. in the rest of the world.
So while we haven't expanded quite yet to tap into the international markets, there's probably an additional $1 billion of opportunity, pipeline opportunity in the rest of the world that would demand the use of this same technology. The underpinning for us in the defense world is a concept called Joint All-Domain Command and Control and Ground, Airborne, and Maritime Intelligence Surveillance and Reconnaissance. That's a lot of fancy words that consultants get paid a lot of money for that basically mean any aircraft, tank, vehicle, submarine, ship, anything that processes sensor data or manages targeting or execution of weapons platforms is managed in that. And Joint All-Domain Command and Control means all sensors, all people, all weapons, all controlled by all people at all times.
So you can imagine anything in a battlefield trying to link everybody, everything, and all processing together, share it all, process it all, is going to take levels of autonomy and artificial intelligence to do that. So the underpinning of this is really the compute capability that our company brings and why, as I mentioned earlier, you need to move from the yellow line to the green line, which is what we offer. On the commercial side, we're finding early applications in commercial aerospace. As I mentioned, autonomy, autonomous long-haul trucking, buses, vehicles, cars.
And then we're doing some things also in this small data center area called composable infrastructure, which is an off-data center capability with an intense area of number of GPUs per square volume that allows people to use CONEX boxes or establish data centers out at remote places or off of a main data center and be able to use a lot of GPUs to go do that. It usually requires management of power and management of cooling and dense spaces, which is part of the ruggedization of the technologies that we do and we provide. This is our Bressner company. I mentioned it's a value-added reseller. The company has generally had a growing profile on just value-added reselling. They've also served as a channel to market for the company to move our more rugged edge enterprise-class hardware out into the European market space.
Those opportunities I talked about as we're going forward, we've got best and leading-edge, leading-class rugged products. We have standard product lines. Those product lines we're also able to leverage into the defense space. For people who have done any work in the defense space, there's an element called commerciality. If you can leverage a commercial product technology or your IP into the defense space under the Federal Acquisition Regulation, you're not required to release your cost and pricing data, which means you can protect your profit margins going in. Otherwise, you're subject to the profit margins in the Federal Acquisition Regulation, which are usually in the 9%-11% range. So really good reason to have standard product and to be in the commercial market and share that dual-use technology. Our idea really is to build out into these platforms to create and generate incumbent positions.
The idea is once you're on a platform that's out at the edge, customers tend to not move from their architecture and the people who are supplying the compute. They tend to tech refresh that every couple of years. You tend to support it. You have a long-term program. There's a current program we're on today for the US Navy. It's a P-8 aircraft. It's an airborne sensor aircraft. We've been on the aircraft for five years. It started with a $2 million program back in 2018. We're now up over $40 million of revenue over that timeframe on that. We've just signed another five-year extension program to continue to support and tech refresh that aircraft. So you can see there, right, 10-plus years we'll be on that platform continually generating revenue every year.
So our strategy is built on continuing to find platforms like that and stacking those up over time for continued and sustainable revenue and profitability. And then I mentioned it, we'll be transitioning in 2025 to starting to tap into international markets. I've had a broad career in doing work internationally, so we'll expand the company's capabilities to take our products and solutions internationally and probably double the size of that pipeline. What do these products look like? So here's a look at the top. We do everything from servers, storage capability. As you can imagine, if you're doing AI/ML, not only do you want to process that capability, but if you can store the data from the sensors or systems that you're operating on, that allows you to bring data back to the large language model or the AI learning element and you continue to process.
A lot of our customers not only demand the high levels of enterprise compute, but the ability for large storage volumes. Generally, in that yellow line of technology I talked about, they use Ethernet as the backplane to link processing chipsets or cards together or other boxes. While we can do Ethernet, we generally tend to use PCIe, which is another open standard used in data centers. What that allows you to do is achieve latencies 1,000 times faster than Ethernet, which can be important, as you can imagine, on an autonomous vehicle, inputting controls into that vehicle, right? You want to have the fastest amount of time from the compute to the actuation of inputs for safety and for operation. That composable infrastructure is built on stuff here, what we call scale-out or expansion.
This is where we're able to build off a single server and stack up boxes of GPUs that provide massive compute capability to a data center, a company, a university, a research center, or an operational unit in the military out in the field, and then we do augment that with some software. We do have to do some software capability built around our storage units and how to best manage and move information and store it, and then we also have some software that deals with the management of the processors or the servers, fan speed, throttling GPUs, monitoring heat and temperature inside of a box, updating and loading firmware, so we do have some management control on those areas. This is what I talked about growing the pipeline.
We've essentially, in the 18 months I've been at the company, we've kind of taken it from $850 million to $1 billion. And as I mentioned, as we add in commercial, we'll be growing that up over $2 billion. You can see it's spread out fairly evenly across those ISR areas and aerospace. But even more so, I think I would highlight over here is if you look at this 62% multi-year new programs and new customers and multi-year opportunities, this represents about 90% of our pipeline is built on those kind of opportunities where we're able to find a way into a platform that would be recurring revenue and we can establish an incumbent position. So we're not just flying customer to customer looking for award to award.
Both in commercial and defense markets, we've identified opportunities where if we land in a platform, we can be there for a long time. The other additional benefit to landing in those multi-year programs is you have the opportunity to expand. If we secure the storage on something, the next thing we look for in that architecture is to get to compute. If we get to compute, we look to add the storage. And then we look to build out that architecture for more opportunity within a customer that we've sustained. This picture just goes into that same story. You can see the concept here of each of these lines represents a pipeline or a multi-year program or contract with a customer. Our idea is to stack those up over time, and then each of those continues to generate revenue like an annuity year on year on year.
Currently, inside of that pipeline, there's about $450 million of opportunity sitting in those kinds of opportunities and programs. A lot of this is preceded by getting customer-funded development where we start with a current commercial product and we will augment it, tweak it, or develop it further down the line for a bespoke application. This year, we will have gone from essentially zero to about $1.8 million in customer-funded programs, which lead to these kind of positions on platforms, and we anticipate by the time we get to 2028, we'll be doing about $12 million a year in lead-in defense or lead-in commercial or defense development programs that will lead to production and sustainment in the long run. This is really just a summary of those elements I talked about. The thesis really in the winning proposition of the company is we've got this resilient global market, right?
We're driven by an architecture change that has to happen in the market because of AI/ML and autonomy. That is not going away. We're uniquely positioned for that to happen. And it won't be a one or two-year change. It will be a sustained element change for decades to decades to come. We've got differentiated products and capabilities. As I mentioned, we're in that top right unique corner. There's some interesting things we've learned in the couple of decades we've been doing this work on how to use the data center hardware and software to work at the edge. The data center stuff was not meant to do that. But the way we ruggedize it, the way we work the firmware, the way we build up the board structures in the processing boards is a unique manner to us in how we make that happen.
We have some patented technology on how to transition people from the compute on that yellow line I showed you to the green line enterprise capability that we do. These platform positions, I think, are a real reason for the foundational success of the company going forward. Build a strong, steady revenue foundation that you can count on year on year that ensures the company has good cash flow. It has revenue. It can maintain its headcount. It can maintain its technology. It can maintain its investment year on year in new products and new capabilities. Establishing those incumbency positions and moving ahead becomes really important. As I mentioned, the pipeline that we've built and developed and that we're prosecuting right now really supports that kind of growth.
That growth for us domestically right now can be on the order of 25% a year given that pipeline we have. The margins in that area can be on the order of 40% or more. That'll lead us to double-digit growth as the company continues in the years to come here. I'm just going to highlight one thing on the capital allocation. While we design our own boards that go into the servers, etc., we generally outsource those for build, and then we bring them back inside for assembly, test, and then delivery to customers. We're not capital intensive from that space and operations. We actually have enough room and space right now. We could more than triple our revenue without any other capital investments into the operations and manufacturing of the company other than adding headcount or extra shifts to be able to process the added revenue.
We're in a really good spot from a capital perspective to drive the company forward. So kind of closing, we'll turn to some questions. Our priorities are to get that growth moving by prosecuting the pipeline, continue to move our gross margins forward. If you've read some of our past statements, we've moved on from some legacy media company business where we were essentially a contract manufacturer. We've now replaced that business with higher margin defense and commercial work, so we're excited about that. Continue to manage our expenses while investing in R&D, and we'll continue to maintain a strong balance sheet, which we feel like we're in a really good position as a microcap company. The rest is just 2023 financials.
You can read about those in the last two earnings releases, or on November 6th you can hear about our Q3 financials and check out the good things we've been doing. My fine friends in the back told me I had five minutes, and that's what I was striving for for questions. With that, I'll open it up to any questions. The stock ticker is OSS, so you can jump in any time and invest.
That seems like a long shot because I'm trying to think of an application in enterprise, let's say Home Depot or something like that. Something that's non-defense area. Is there any interest in there? Is there any opportunity there? Because I'm not familiar with that.
Yeah, something like a Home Depot, unless you're in a manufacturing space, right? We're using a lot of sensors to automate a manufacturing line.
Or if they really wanted to build a data center of their own in a specific area off like a headquarters and they wanted to have a GPU concentration, then they would find us. Short of that, right, then they're better off through the cloud into the data center or just standing up a data center inside their own building.
So going to enterprise, it's going to be exactly a focus in the near future at all.
Yeah, yes. Yeah, for us, moving to something like NVIDIA or Dell, that kind of top left corner of that graph, it's a competitive space, not for us, not for our where our differentiators are in technology.
Most military that.
Yeah. Well, military and commercial, right? Anything out on the edge.
When I use enterprise, right, I use enterprise class to mean taking that data center capability, right, and delivering it to the edge. Example on the commercial side, we do some stuff in commercial aero on commercial airlines. We do some stuff today with Andretti Racing. If you watch their Indy 500, we do all the processing for their trackside AI analytics.
To dovetail on that, in terms of the penetration in that available to the market for enterprise, what would you say percentage-wise, roughly speaking, that you've already made?
Yeah, the question was, what might our market percentage penetration be in that commercial side? It'd be very low. I mean, we're a $60 million company, 30 of which is in this space. I mean, we have lots of room to grow to take market share. Yes.
Seems to be a lot of operating expenses this past year, lots of loss in the operations.
Yeah, so the question was, there's a lot of operating losses here in the financials. That was really driven by two things I kind of touched on a little bit here. The first was we moved away from our media and entertainment business. That was just before I had joined the company. So the company had not contemplated before that without any experience really in the defense market, how long it takes to replace revenue with defense revenue. So that part has taken a little bit longer to recover from. That same part. The second thing was, the company and the board, we decided to refresh across a number of our product lines to reestablish some key positions as especially as we wanted to penetrate further some of these markets.
So we had a little bit higher, I would say, R&D spend than we had in the past years to facilitate us being ready. And in part, that was the defense procurement cycle can oftentimes be more than a year. It can be two years. So the idea was to get products readily available so we could try to accelerate that and pre-position ourselves. So we started that clock sooner where if we had drug out those investments, we would have just been dragging out the timeframe for getting back to this growth. And if you've watched the last two earnings releases, you will have seen not only have we grown revenue over the last two quarters, but our book-to-bill ratios have been about 1.25 or higher. So while the company's going revenue-wise, right, we've been expanding our bookings by 25%, which goes back to everything I was saying.
We believe the company can continue to grow organically at 25%, and then we'll reverse what you've seen from the operating losses.
There are a lot of bullet points on there saying, describing continuing R&D to, I guess, develop product leadership incumbency. Can you go into that a little bit more?
Yeah, so the question was, we had some bullets on product leadership and incumbency, so you'll see in a lot of our spaces, right, especially in the rugged space, right, we really get the enterprise, the data center stuff down into tightly packaged areas, and we build those standard products so they're readily available for the market. In that top corner, right, we do that so we maintain the leadership position in products.
So in that capability, like we're probably the only person I know of in the U.S. and serving the military especially that has that kind of rugged capability at that level with standard products. The other thing it does is influencing the transition timing from that yellow line to the green line I talked about. We will use our products and demonstrations to demonstrate to customers why they should continue or in their next tech refresh why they need to change architectures. I got 15 seconds, so it can be a really quick question or a quick investment in the company. All right. Well, thanks everybody. Enjoy the rest of the conference. I appreciate it.