All right. I'd like to introduce our next presentation here at Planet MicroCap Las Vegas, powered by MicroCapClub. We have Bill Wentworth from Data I/O.
Thank you. Thanks, everyone. Thanks for spending the time. You guys enjoying the nice cool temperatures outside? I'm a hockey player and skier. Trust me, this is not fun. Just the walk over was a stress. Going to spend a little bit of time. We've got quite a few announcements we've made over the last, since May 14th, things that we've been working on for a while. I became CEO, let's see, a year ago. Call it January. I was there for November and December, that was kind of discovery and kind of getting your feet wet and try to get an understanding of really what was going on in the company. We've gone through quite a transformation. There's a lot of detail and content here. I'll wait till the end for questions, I think you'll see what we've done, pretty interesting.
Our safe harbor, I won't go through that. You guys see that all the time. Data I/O's been around since 1972. I was actually ironically introduced to this company in my early teen years. My dad was in the semiconductor industry as a distributor, eventually sold his company to Arrow. I grew up in this space. I worked on the very first piece of equipment Data I/O ever made, called the Data I/O 1. That's when actually a master was like a punch tape, and you would push in binary code, and it would push out a piece of paper tape, and that was your file. I've been in this industry for a while and started my own services company when I was 21, using Data I/O as my only platform. Back then, Data I/O had about 80%-90% market share.
They were the leader by far. Over that time, they kind of lost their way as far as automation and solutions and things as programmable technology became more proficient and more, certainly, across a lot more products. As those volumes increased in this technology, there were other players that came on board, such as BPM Microsystems, System General, and other competitors, which really started to challenge Data I/O over the years. The background of the company, the one thing that really attracted me to take the role, was the tenure of the employees, the talent that we have inside the building, been there for 30, 35 years. This is a very deeply experienced technical engineering group of people that really understand this technology, that have been around longer than most of our competitors. We're in a very niche part of the semiconductor industry.
Like I said, programmable technology, flash microcontrollers, flash memory. It's about 20% of the overall semiconductor content that's sold today as far as unit volumes. What it is, our equipment basically takes a file from an OEM, say Apple, that's why we use a phone, and that software file gets wrapped by our software and put into the semiconductor part through our platform, which is the product in the middle. That hardware and software combined is what takes that pattern and puts the IP, which is what these OEMs are giving us, are giving our customers, that IP to put inside the silicon to make the product work. It allows that finished product to do its final functions and its technology, and actually brings it to life. These are some of our customers.
As you can see, there's a lot of big names up there, Fortune 500. We've been dealing with a lot of these customers for 10, 20 years, so longstanding relationships. This is the management team. As you can see, myself came in. I've been in this space for about 30, 35 years. Monty Reagan used to run sales for BPM Microsystems, one of Data I/O's larger competitors. I brought him in the first week I started. Charlie just came in as a new CFO, a lot of experience in enterprise software. Excellent CFO. We've got a great partnership. He's done a great job of cleaning things up. Duane, like I said, 35 years with Data I/O, ton of experience. Cindy brought in as our new people officer, and Doug Adams is my old COO from Source.
We're going to be bringing him in to run our new organic services strategy, which I'll talk about a little bit later. This was the portfolio I inherited. As you can see, these products are a little antiquated. They didn't hook up to hosts, so there was really no product that could go into the engineering departments of customers to expand the customer base. This is kind of where Data I/O really fell off. They focused too much on automation but didn't focus on the core. One of the first things we started to do was invest in the core business. Performance-wise, over the last five years before I took over, a lot of cash erosion investments that didn't really go anywhere.
The last year and a half has been really a transformation year, really investing in the new platform, invested in the core business, the core LumenX platform, which was really a product at the time when we came in. We've right-sized the operational expenses of the business. We've taken about 22% of cost out in the last 12 months, most of it in the last six months, modernized the IT, and we launched an M&A strategy in Q3 of 2025. We've got a unified programming platform now that spans across a larger group of products. We've opened up our markets. Data I/O was mostly focused on just the CapEx part of data provisioning, which was only a $130 million market. We've launched the platform into services and also at test, which opens up to about a billion-dollar market.
This is the transformation kind of roadmap and what we started, joined in November, formed a new leadership team, invested in the core platform, really changed the business from front to back. It just needed a lot of work. Hired Charlie s tarted in September. We announced our new unified platform, which was completed in Q4 2025, restructured a couple of our operating units in both Germany and in Redmond, and really have put AI to work quite a bit. I'll answer any questions about that later, but it's been a huge help in reducing costs, but also optimizing the business. As you can see, it's a pretty big transformation of products if you compare to the last slide. All those products hook up to hosts. We get them in engineering systems, in engineering departments, and it's really starting to grow our new account base.
We've landed six new logos in the last quarter alone. This is our first kind of change in our go to market. On the left-hand side is what we serve today, which is the CapEx market. Again, a small market, about $130 million. What we're converting into is Programming- as- a- Service. Taking our existing platform and providing services as opposed to selling CapEx. Better margins, better cash flow, stickier multi-year contracts. It's what I used to do in my old business at Source. It's something I'm very, very familiar with. This is what we announced on May 14th, our last earnings call. We just had a $9 million placement from a private investor. We actually just got funded today.
Announced a $23 million acquisition on May 14th as well, which will give us scale and scope and expand our markets from $1 billion to almost $100 billion. We've been mostly focused on automotive. This takes us outside of automotive and expands into multiple new markets that we've never served before. We expect significant growth in the out quarters of this year. We should close the transaction sometime the end of July, and that'll be our first quarter that we can announce the earnings of both companies. Commitment to delivering shareholder value, it's one of the things that this company has not done over the last 5-1 0 years. I can tell you we're aligning our compensation to shareholder value, meaning we get compensated as the share value goes up, which is kind of the first thing this company has never done.
Expanding our TAM 10x with the acquisition. Improving our revenues, doubling down on what's working on the platform. The transformation acquisition that we announced on May 14th will accelerate this plan and give us scale and scope to do that. I think that's it.
[audio distortion].
Yeah.
[audio distortion].
Yeah
[audio distortion].
Charge a piece price. Yeah. There'll be a managed service fee that will include hardware, software support, the cost of the investment of the actual equipment will be spread out over a three-year contract at a cost, and then we charge a piece price.
[audio distortion].
Yeah. We'll still own it.
[audio distortion].
It'll be on-site. Yeah.
The good thing is we have 650 pieces of equipment installed already. A lot of times people think that, Oh, I buy a piece of equipment, I turn it on, everything's fine. It's a manufacturing process, and they find that it's a lot harder than they thought. A lot of them want to go to an OpEx model.
[audio distortion].
Yeah. We take the labor. Yeah. We'll train them. They could go on our payroll. They could bill us the labor. There's a lot of different ways you can manage that. Yeah, we would own the labor, the equipment, the support, the engineering support, all of it. Yeah.
[audio distortion].
Yes.
[audio distortion].
That's the point. I can go director level here, right? It puts more control in their hands.
We did a two-week stint in Mexico and got $7 million of pipeline annual in two weeks. We just launched it a month ago. We will be a software services company. Still sell CapEx, but primarily we'll be software services.
[audio distortion].
Yeah. We'll give them value back. Exactly. Yeah.
[audio distortion].
Yeah. You can sell better than my salespeople.
[audio distortion]
Yes, they do have a services component, and it does get us a regional programming center, which accelerates this organic strategy because you need to have regional too. You have small customers that don't have enough volume. Yeah.
[audio distortion].
Excuse me?
[audio distortion].
We're pretty halfway through due diligence at this point. I don't see any risks associated. It's a pretty clean company.
[audio distortion].
They didn't. I went after them. Who wants to wait for an auction, right?
Look at it, the markets are in it. It's a company we've known for a while. I've known for a while. It's an adjacent play directly. You know everyone. It's not hard when you're in the small part of the semiconductor industry. Yeah, I know all 10 of them. Yeah.
[audio distortion].
Yeah, interesting. When it really started to hit news lines a couple of years ago, as a board member, sponsored an AI project to just read documentation, because when we get specs from a semiconductor manufacturer, they're 400 or 500 pages long, and you're looking for 10 values. We decided to create a little doc control AI that would read this spec, and it would normally take our engineers probably five days to read it to get the values. We got it down to about three minutes.
Since that point, it's now far more automated, far cheaper. Back then, we had to actually get our own GPU machine to do it. That project cost about $130,000 to do, but it saved a ton of time. We did that same project in about three minutes, and it cost us maybe $0.48 because of just loading it into the models. We've used it for software development. We've got code start to finish being managed by AI and being released. We've used it to, people that leave, software developers, we've replaced it with just code agents. It's helped us out a lot, a lot of different ways. Even customer-facing too.
[audio distortion].
Yes. Yeah.
[audio distortion].
Yeah.
[audio distortion].
You still are.
[audio distortion].
Yeah
[audio distortion].
Yeah
[audio distortion].
Yeah. Yep.
[audio distortion].
Yeah.
[audio distortion].
It gets too big.
[audio distortion]
Yeah.
[audio distortion].
Well, I've seen it like-
[audio distortion].
Yeah.
[audio distortion].
No, I think that the AI, as far as the software code that our customers are doing inside these components, have limits. There's only so much memory they can fill.
[audio distortion].
Yeah
[audio distortion].
They could, yeah.
[audio distortion].
That's more for the in-memory of the AI and the models and the tokens and stuff that are sitting there. That's why that stuff's taking off. It's just so they can get more in memory so they can get answers back faster. For us, it's the memory sizes that the IP goes in probably won't reach over a one terabyte file size, so it's not huge. That's a fixed pattern to do a very fixed function. Like microcontrollers have a small microprocessor in it, and then there's memory, and that memory gives it some added instructions. But things like robotics, we're working on a project right now as a humanoid robot, in the shoulder has one board, has 22 microcontrollers on it. You build 1 million of those, there's two shoulders. You can do the math. As a service provider, that's a huge opportunity.
What takes time? [audio distortion].
It's transferring the data onto the chip. Yeah. When you get into UFS flash, which is high dense, almost like a mini hard drive, those can take up to 5-10 minutes to load per part. That's why we have larger machines that can hold multiple test sites. You just have to do that through scale and speed. Yep. Yeah, those are probably the longest programming times. Micros take 5-30 seconds. Yeah, pretty quick. Any other questions? All right.