Micro Systemation AB (publ) ($MSABB)

Right. So good morning, and welcome to MSAB's Capital Markets Day. I'm Carolina Marteus, and I will be moderating today's session. So, we're here today to get a better understanding of the company, where it stands, what's driving the business and how management sees the opportunities ahead. Before the break, we'll hear from the CEO and CFO on strategy and the financials, and we will also open up for some questions. And after the break, we go deeper into the market and the products, and then we move into a panel discussion on AI. So, with that, let's go. Peter Gille, the floor is yours.

Okay. So, my name is Peter Gille. I've been the CEO of MSAB now for almost two years. Before that, I was in the Board. It's been some exciting times these 22 months or so. And today, I want to tell you a little bit about how I see MSAB and our future. So, I hope you will find it interesting. I'm sure you will find it interesting. So, let's start with a little bit about MSAB. What is it? It's a small company. We're around 200 people. But still, we are #3 in the world in our niche. So, we are -- it's a high technology niche with high barriers of entry. There're very few companies in the world that can do what we can do. That's the reason we serve customers in over 100 countries. 200 people serve customers in over 100 countries. We cover over 50,000 devices. In our office, we have 15,000 physical mobile phones that we can go down and pick up. Some models that I promise you, you have never seen before. So that is really this small company with a high technology expertise. But what do we really do for our customers? What we do is really that we pull information out of these devices. Many of our customer cases we can't speak about. There's a lot of test about MSAB, they don't communicate. And the reason is that in some cases, we can't communicate. We help terrorist fighting units to prevent terrorist crimes. We've done that in several cases in different countries. But we also help the police, for instance, getting the data out of burner phones where they have drug dealer lists and things like that, as well as we help the FBI monitoring how the guy who tried to murder Trump moved around. So, there is a variation of cases. We have border control, check people that comes to the border, if they are terrorists or check that their story that is the correct one by taking data out of the mobile phones. So that's really what we do for our customers. And to do that, we need to use vulnerabilities in the mobile devices. Exploits, which we use to get into the phone and get the data because they are usually locked and the criminal, the drug dealer, he doesn't want to give this away freely. So, the police need to use our technology to really get the data. And that's something that we use within MSAB, get the data. That's what our customers want to do, get the data and analyze it. So how did it start? It was a long time ago. It started really with the birth of the mobile phone. There was this small company doing cables, working with mobile phones. And then they got a question from the police in Sweden, where a police officer needed to get a telephone list out with calls. It wasn't a lot of data in the mobile phones at that time. But he wanted to solve a case, and that could help him solve the case. So, with the technology that MSAB had, they could utilize that and get the data out and the officer could solve the case. And of course, there was -- maybe this could be used by other police forces, maybe there was an interest. So that's how it started. And MSAB was really first out on this industry and been in the industry so far and being quite successful but had a very typical Swedish approach by being careful, making sure that you grow steadily, slowly but always making a profit. At that time, competitors also discovered this niche with bigger resources, more aggressive approach and some of them grow bigger than MSAB. I think that was a mistake that MSAB was so careful. We're not careful anymore. We are here to grow. We're here to invest. We want to make profit as well, but it's not our #1 priority. So, you are here as investors. Let me give you some reasons to why I think you should invest in MSAB because I really think you should do that. Most of you have done that already, but you can invest more. When I joined, we defined really a strategy for how to grow. And in my first report, I said we have defined a strategy. In essence, it's invested in the products. We changed the way the company works. And then we have executed to that strategy. And we said also at that time, after 12 to 24 months, we will see the results. We now start to see the results already. And in the 3 last quarters, we have grown double-digit numbers in local currencies. So, we have a strategy that works. That's a good reason, maybe the most important reason. Also, one of the things that I like most with MSAB is that it's so scalable. I love scalable businesses. And most investors do that as well. We have a high 90% gross margin. We don't need a lot more staff if we sell a lot more products. It's really scalable. Even if we invest, we will have stairs when we increase the cost. But in essence, we are a really scalable company. So that's the second good reason. The third is we have a really good market position. We are #3 in the world, but we are also the only European player. And Europe starts to become more important nowadays, buying European. I hope that will give results as well, especially in the important sectors like military and other defense organizations. So, being #3 in the world is most companies dream about that. We are already there. Also, our products, we have really invested heavily in our products, and they are today world-class products. They are really high class. It's the best products in the market. It goes a little bit up and down depending on how the patches from the mobile vendors apply. But in -- we are on the top now. We will hear more about that later on when Tomas comes in. But we have really, really good products. And this is a product business. Our customers, they buy the things that can get the data. So, if you can't get the data, if you are -- you cover only 75% of the phones or the mobile devices, they will not use your solution. So, you need to be there at the top. And that's one of the things we changed. Then we come to the organization. We have done a total makeover of the organization. The management team, I think Tony now is the most senior guy in the management team, and he's been in the company for 3 years, and he's done a fantastic job. Many are younger than me. All are younger than me in age but have been with the company less time than me. So, we really have a totally different management team. Before, I think we had the management team's been there a long time. They got stuck no matter how good you are. After a while, if you've been there for 10 years, you get stuck in your different patterns. So, you need change. Okay. We've done a lot of changes now. but it seems to be working really well. But we haven't only changed the management team. We have also changed a lot of other people in the organization, and we have also changed how we work, how we do processes. And I think we have a really, really good organization working now. It's the best product and R&D organization I've ever seen -- ever seen in the companies I've been to. And that's an important part. But it's also starting to be a better and better sales organization. Marten will talk a little bit more about that. He is new, but he's already doing changes that's needed to the organization. So that is five reasons to invest more in MSAB or invest in MSAB. If you look on the market, it's really also -- it's growing at least the needs. We have a lot of more data. The complexity of the device is increasing all the time. It means that the barriers of entry for new players to coming in is higher and higher and higher. It also means that we, as existing players, have more challenges. There's a lot of regulatory changes that our customers need to adapt to. The geopolitical things that is happening is also affecting this industry. And then we have AI. I think we will see both a positive and a negative impact of AI for us in this industry. I think, in essence, we are more protected than a normal software company because we have a lot of hardware solutions here or close to hardware solutions. But -- and we can have the fruits of developing faster with AI and so forth. But it's going to have an impact. That's for sure. Mostly, I believe, positive impact. Looking at the market, it's a market that doesn't change a lot. We have the biggest player here, an Israeli company, Cellebrite, fairly aggressive. They are the biggest player in the market. They are American, but everybody are from Israel in the company. So, we say they are Israeli. And we have Magnet, a North American, Canadian company, which we're partnering with, but we are also competitors in some cases. They are more specialized on the U.S. And then we have MSAB, being smaller, having a smaller market share, but are growing. And then we have the rest. We are, as I said before, the only European player in this aspect. So, I think that's a good thing to keep in mind. Who are our customers? What is the industries that we serve? We serve law enforcement, as I said, the police organizations, various law enforcement. It's about 50% of our business. It's a big part. Custom and border, utilize our solutions, utilize our frontline solutions, sits at the airport, checks people that comes without the passport, but you have a mobile phone, okay, let's check if you really are the one you say that you are and if you are a terrorist and so forth. Military and defense, it's super interesting. We serve the military police. We serve special ops and so forth with our solutions. Government related is our second biggest. That's where we have all the 3-letter agencies that use our stuff, but we also have a lot of other government facilities that use our products. And then we have the private sector, very small for us. We are being careful in the private sector because we are under the -- we need to act in according to the European or the American act how and who we can sell to, the regulations. And if you sell to a private company, you never know where they will use the product or end up. So, we are very careful with the private sector today. The users, they are investigators and analysts. They are using our software. It can be forensic examiners digging deep into the technology. It can be frontline personnel in the police station. You can take a witness statement out from the phone. You can be at the airport in the border control. Well, being in the Frontline. It's a kind of a unique offering we have also there. And then it says managers, but it's more the IT tech people that are setting up the solution, how it should work and so forth. So, it's more the IT manager side that use it. But let's see how it looks for a real customer, what they are doing for us. Should come a movie here. So, while they are fixing this. So, this is really about the case in Nevada. It's about child exploitation. We work with CRC, which is the Child Rescue Coalition, which is an organization that really focus on exploited children. And we work with them as a part of our sustainability and social response. So, we give them our solutions to really help the police in various cases around. It can be things in the Philippines or in other areas. This is more a normal case, but this is a typical case where police use our solutions. Here it comes. [Presentation]

So, in this case, they caught the perpetrator. Another example is the long-term customer for us, the U.K. counterterrorist policing, been using our products for a long time and have prevented a lot of bad things happening with them. So that are two examples, but you will hear from a real customer here today. And I know that I will tell you more. Our ecosystem is really -- it looks like this. We have our products, our software, to that, we have a number of services and then we have the hardware aligned to that if you have a frontline solution, depending on how you use it. And then we work with a number of selective technology partners because in some cases, you want to have information from a vehicle, then we work with Berla. You want to have it from a PC; we work with Detego. If you want to have it from a drone, we work with V2. It's a very similar technology to us, and we can analyze all the stuff, but we need to use the partner technology. Okay. Last but not least, how are we going to grow? What do we see the way we're going to grow going forward? First of all, we see now that we are winning share in the existing markets. In the mature markets, this is really the game. All the customers in the Western world, you would say, they have bought the solutions normally from all the major three vendors. But we are the smallest one, and we are now growing on each of these customers by providing the best product in the market. And we see that that's happening. So that is one growth avenue we see if you look on a geographical space. But we also think emerging markets, for instance, in APAC, in Asia, is super interesting for us. They have more Android devices, which is really our core. And it's not -- it hasn't established itself that market yet. So, it's a much bigger opportunity for doing national-wide deals, doing bigger deals and establish yourself. They are more open to look at different solutions right now. So, we think that will be a growth market for us going forward. And it's something we're really focusing on right now. Then we have military and defense. What we have sold historically to the military side has been -- normally, we sell to the military police, or we sell to special unit forces, which go into a country and grab the phone and check it. But what we see in Ukraine, is that they are using our tools really as an offensive weapon in the field. So, they grab a soldier's phone, they check in the Frontline, okay, where did he sleep last night? How has he moved, what messages has he sent and then they can shoot in that direction. So, it becomes a more -- much more offensive weapon. To utilize that even better, we need to do some more investment and development. But we think that this would be a really interesting segment going forward. And it's a lot more licenses to sell if you sell to the whole army rather than to the military police and the special forces. Private. We think that private can be interesting, but only in the way that we want to help identify malware on the phones. If it's been infected, you've been traveling, you want to check your phone. We think that can be an interesting market. That's something we're looking on. So, from a product side, these two solutions, an easy-to-deploy triage solution for the Frontline that also goes into the military and the malware detection side that goes into the private. So that's really where we see that growth will come from growth avenues in the future. We are already growing, but this is for the future. So, with that, I will give you what you really want to see the numbers from Tony, our CFO. Thank you.

Thank you, Peter. So, before starting off then with this morning's news, the financial goals for the company, I'm Tony Forsgren. I'm the CFO of the company since a little bit over 3 years. Yes, this morning, we released some financial goals. Our long-term targets, and they are to serve as guiding targets for the company and our strategic direction through 2030. By year 2030, we are aiming to reach SEK 1 billion, an EBITDA of at least SEK 200 million and last and not least, a dividend target of between 25% to 50% of the net result. The long-term financial goals are forward-looking statements and should not be taken as forecasts. I know you are sitting and calculating already. That said, let's have a look at our financial situation. But before that, also a little bit of a history. I'm going to present more a forward-looking slide show here but also talk about a little bit of the history, which Peter touched upon. This is our first CMD. We are proud to have it, and we feel that we have a stable platform to have it. We didn't have that platform 3 years ago, but we have it today, and we are confident in the numbers we are showing internally and externally to the market. We know that MSAB has been hard to follow as a company. And we have made huge changes, both in personnel, management, strategic direction, also our systems, ERP systems, product portfolio, price segments, terms and conditions with customers, everything I can tell you. It's been a hard work, but we have a stable platform for growth now. The aim is also to improve our communication with the market. Let's start off with the revenue. This is a tricky one for some. So, I'm going to spend some time on explaining how we take the revenue. We sell perpetual licenses. Previously, that was a long -- they didn't -- there wasn't an end to the license. It was really perpetual. We have changed that in our terms and conditions. So XRY Pro, our most advanced product, is ending day 1 if they don't pay the new license. So that is not really perpetual. And our biggest product, Logical-Physical, it ends after 12 months of the expiry. So, it's not really perpetual anymore. And that's also gaining the revenues to us. So, you can see to the left here, the pie chart, the perpetual and the model we have today, we take 84% of the revenue day 1. And the rest of the 16% is distributed evenly over the contract length and that stands for support, service and updates of the software. The 100% cash upfront is really good for our cash flow. But it gets -- if you look to the right, it gets really volatile between quarter-to-quarter in what numbers we present to the market. And as you can see in the dark blue colors, it's the half year 1. The light blue is half year 2. We really do better in half year 2, and that follows our customers' budget processes. But also, if you look at the status between quarter 1 and quarter 1, it gets bumpy. So, with that said, we have reliable customers. We have strong cash flow. And I also want to tell you, we have a really, really low churn. I will come back to that. So, what we have done then and for the first time presented in this Q1 reported for 2026, we have an internal KPI called annual contract value. We have been following that internally for a while to see it, to see the growth, to see stabilization of it, et cetera, and talked internally, are we going to release this to the market. And we feel confident that we did this Q1 report. So, what is this annual contract value and how do we measure it and how do we calculate it? You can see up to the right again then how we're taking revenue today, 84:16. And if -- like in this example, having a 1-year contract and extracting the numbers to the annual contract value base that we count, we split it evenly over the 12 months. So, for each and every contract, this is not real numbers. I can tell you that. This is an example, it's illustration. So, for each and every contract that we have, we pull it line by line, year-by-year and distribute the revenue in even parts over the contract length. And it's a lot of contracts; I can tell you that. And what did we do then now in March in the example I'm showing January? What do we do? We take a snapshot in March, what is the revenue for every contract that are active this month, and we multiply it by 12. So, we get the annual value of the active contract base that we have today. And then we look back 1 year and see what was the value 1 year ago. And that was what we are presenting in the Q1 report now. This is really following, in some cases, the revenue recognition growth, but not -- it's not 1:1, and I will tell you why. But as you can see, we grew 38% March this year versus March last year. And you can also see we have been flat, and that has also been reflected in our reported revenues for -- if you exclude 2025 last year, which was a real successful year. Look back, it's ForEx and FX and stuff like that, but we have been flat for like 5, 6, 7 years due to no investments in the company. So, this -- finally, this is not to be compared with what we are reporting in our external report as recognized revenues, and it cannot be bridged. I know we're going to calculate when you come back and try to bridge it to our reports. But I can tell you, we have all the facts and the data about contract lengths, the volumes, the price mixes, the FX, the discounts, et cetera. We have all the data and have ourselves tried to bridge it. We cannot do it. So, we have to see it as a snapshot of the annual contract value today. Okay. Since we are selling perpetual licenses, new sales are very important for us. I will give you an example now again. So, if we have a contract of 100 for a new sale year 1, the second year is approximately 55% to 65% of the total value for year 1. So, we really have to fill the bucket with new sales. And we do that. And we measure and monitor this KPI very closely because we have to bring in more new sales to our business. And we do that, and that is showing in the growth. So, our goal is to be at least 50% new sales of the total sales and now we are there. We haven't been there. We see '24 wasn't that good, and that showed up in the reported numbers as well, '25 is really good. And rolling 12, '26, we are still there, 51% in new sales. So, as I said, we are following this really close. What is it driven by? It's mainly driven by our product investments and mainly with our outstanding product, XRY Pro. Okay. I'm following the P&L going down from revenues, and now we are at gross margin. We have a stable and we have stabilized our gross margin, and that is a key driver for scaling our business. Hardware represents a really small cost in our P&L, and it's based and attached to new sales. In the past, we have low margins, unstable gross margins, but that was due to pure hardware business with customers. We don't do that anymore. We have stopped that. And the seller, to be really honest, they got bonus for selling those hardware. They don't. So, we don't incentivize it at all. So, we are just scaling and cleaning and want to be a pure software company with a high gross margin scalability. Okay. Costs, OpEx. 77% of our OpEx is personnel costs. And as Peter said, it's approximately just above 200 personnel. So, this really sets a good foundation for not scaling the cost in the same pattern that we scale the growth or the sales. We can do more sales. It doesn't drive -- it's no connection, no direct connection to the costs. Maybe if we sell more, we have to have some more support personnel and stuff like that, but it's not driving in the same pattern as we grow with the revenue. So, we invested in our products. You can -- you saw that for the first time last year in the balance sheet. We invested SEK 30 million. We bought exploit capabilities to the software externally, which we balanced. We still have an agreement with that vendor. But now, we are buying support and updates on the software that we bought last year, and that is taken in OpEx. We explained that in the quarterly report. It cost us SEK 4.7 million quarterly, so close to SEK 20 million in support as well as we sell support, the 16% we distribute and updates, they sell support to us to maintain the capabilities with our iOS features in XRY Pro which is really important and really XRY drive the growth. So that said, 35% of the total OpEx consists of R&D. And even though we cover all internal R&D costs upfront, we can keep OpEx stable. So, I've been through revenues, gross margin, OpEx. Now we're coming to the scalability. Revenue gross margin OpEx summarizes a scalable business model. The platform is established now. We haven't had it before. It's now it's established and it's highly scalable. And according to our strategies, we want to expand in emerging markets, new verticals and gaining market share. Also new sales, as I mentioned before, and upgrades in current licenses. And the base for that is XRY Pro, which has a price point of 4x versus the Logical-Physical. So that said, 2026 it's somewhat an investment year. We will grow in sales, but we will also continue to invest in products and know-how, personnel. However, this is an investment year our EBIT is planned to be at least on par with last year, maybe a little bit growth in EBIT as well, even though it's an investment year. Yes, cash conversion. We have a really strong cash conversion case here. This results in average 86% cash conversion ratio between year '21 to '25. We have a good -- nowadays, we have a good and stabilized DSO, Day Sales Outstanding, which is important for us and the cash. We have a light and debt-free balance sheet. We received payments upfront, resulting in the strong cash flow generation. And investments, as you can see, started off in 2025 and are going to continue. Finally, my last slide, capital allocation. The diagram to the left is just for illustration, so no calculation here. What we did before 2024 was the capital allocation. We have been handing out dividends. We have been handing out dividends to the shareholders in average with 109% between year 2022 and 2024 in payout ratio. And the proposed for the Annual General Meeting next week is 60% of the net result. So annual dividend is brought from '24 and in history. We shifted in '25 to a growth journey. We started to invest. We will keep on investing. And looking this year, '26 and going forward, we will keep on investing. We will keep on giving a dividend to our shareholders. We will also look into M&A. This is where we're going to spend our money going forward. With that said, thank you very much.

Thank you, Tony. Please stay. And Peter, will you please join us for a few questions. Okay. So maybe let's begin with growth in APAC. Asia Pacific is a big focus, but still a smaller part. What gives you confidence you can really scale there?

As I said, it's not really -- it is an emerging market in every way. So, it's an open game, I would say. So, it's more about the best supplier can win a lot of big business. So that's one reason. The other reason is that they are more Android focused. And I think we have -- we are on par with iOS. But when it comes to Android, we are really outstanding. So that's the other thing that makes us win those deals. So, I think that is the main reasons.

All right. But what's the -- I mean, what's the main risk in the execution?

The main risk, I think it's -- those markets are, you never know how the sales processes will end. It's kind of -- it's a market far away with different politics and different ways of doing it. So, I think that is the risk. But we have some really good salespeople and organization down there. So, I think we can -- so far, we've been able to manage that risk.

Okay. And just to make sure we don't miss anything, how do you see the U.S. market? And what's your biggest challenge over there?

I think we have a challenge in the U.S. market as well as others have with the current political situation that's been in the U.S., which has really made it difficult to allocate budgets. I think now that starts to happen. But also, it's been a lot of changes in staff, and you don't even know who to contact. People don't know what they can do. So, I think it's been a turmoil in U.S. the last year. And that is difficult to handle for everybody. I hear this from also the competition. Of course, you can get renewals, but getting new business has been difficult. But we see that, that starts to change now. So, the U.S. market is super important for us. It is a big market for us, but the geopolitical situation needs to kind of stabilize. I mean, when President Trump came, we thought he would invest heavily everywhere where we are important, police. ICE is one of our biggest customers, for instance, on the software side. But he didn't do that. He tripled the ICE budget, but he cut the IT budget in half. So, it's been difficult to predict what will happen. But it starts to stabilize. I think we will see some good developments this year. And we still grow last year in local.

Okay. Good to hear. Let's move on to competition. Let's say your competitors have more scale. How do you catch up?

We don't. I mean it's -- we can't scale as they do because they are bigger than we are. But what we need to do is to focus on what we do better. We are more adaptable. So, we can be much more adaptable as, for instance, with AI, you need to adapt. It's much more difficult to turn around with a big ship than a smaller ship. And so far, also when we look on it, where they -- if you look on the product side, we are really keeping up. We don't see that they produce better stuff than we do, with the organization we have. So, on that side, and that is really key. We are keeping up. Then, of course, they have more salespeople, but they also need to pull in a lot more revenue to keep the position. So, we see we are growing faster than they are right now. And to keep on doing that, we need to be smarter, better and more adaptable.

Okay. Now let's talk a little bit about this year. It's an investment year. So how should investors think about the near-term impact? Exactly. Maybe, Tony.

Yes, it's an investment year. We keep on investing. So, there will be some investments during this year, but maybe you should not exaggerate the investments that we will do. We still think that we will have a healthy profit. We think if we grow the business as planned, we will grow also the profit. So, it's not like we will invest, so we will take the costs to a higher level. But it's not the scalability that I want in the company this year that will come next year.

So, the revenues, will it come -- when will it come?

The revenues will come this year, will also come next year. But this year, we will also grow the cost a little bit more than we will the year after in proportion to the revenues. So, we will scale, but we will not scale as much as we should.

And a little one adding to that. Investments is new for us, this company. So, we started investing. But looking at the numbers of investments we do, it's a small portion to the business to drive and it gives a really good return. That's what we're talking about when we're talking about investments, smart investments.

All right. So, this question maybe is for you, Tony. You're introducing ACV. Just to be clear, what does it tell us that previous metrics didn't? And why is that so important? Could you explain that?

I think it's important for us to describe to the market how the underlying business is going. And that is not told by the revenue recognition we report in the annual -- in the Q1 reports and the annual report. The underlying growth is really on the active contract base, and that is really growing. We saw good efforts and revenues in Q3, Q4 last year. We did almost 15% organically in Q1 '26, even though it's lower numbers than the half year 2. So, we're really growing the base, and that is not reflected in the same way in the revenue recognition since we are taking 84% day 1. If we sign a 3-year contract, we take 84% day 1, and we don't see so much revenue for 3 years. But if we distribute it evenly over the contract length, you can see the underlying contract base that are active and not churn. I'll be back to the churn then. So, I will tell you, we have a churn between 0.5% to 2.5% in value each and every month. So always below 3%.

But how can I, as an investor, use ACV?

Yes, you can have, as a metric, a snapshot on what is the actual contract base today this month.

Okay. But not the future revenues.

No.

Okay. Maybe we start with some questions if we have any questions from the audience. We have a mic. Kim, could you...

So, I have two questions. First one, how long is the average length of the contracts? And the second one is when you're talking about investments, what exactly is it that you're meaning? Is it in staff? Or is it software development seen as CapEx?

So, the average contract length we have been working on, we have been working on shortening the average contract length. It sounds strange. But since we don't have a churn, we want to have shorter contracts to gain the price increase we do on a yearly basis, because if we sell a 3-year contract, we lock down the price for 3 years. If we do it annually for 3 years, we have a price increase in average of 10%. So, we're gaining more revenue. And the annual contract length is 17.6 months today. It's been closer to 2 years before, 17.6 months today. So -- and investments, we are investing in capabilities in our products, which you saw last year in the balance sheet. And now we told it's approximately SEK 20 million in the OpEx for running that service. But we're also investing in personnel this year. So, we are increasing personnel with approximately 15 FTEs this year. And when we've done that in '26, as looking of today, Peter, correct me if I'm wrong, we see we will not do it in the same pace. We will also monitor in AI and see what brings to the organization, and we will scale.

Yes, that's right. And a lot of the investments this year goes into the sales and marketing side as well.

Okay. I think we have some more questions.

Magnet is #2, isn't it?

Yes.

Didn't you say that you partner sometimes with Magnet?

We do.

In what respect?

There are customers that can buy a solution where they use Magnet for, for instance, for iOS, they use us for Android. And so, they use that -- we have a package like that. But some customers I can't give the name on.

No. So they are not strong on Android.

They are capable on Android, but customers think we are stronger.

Okay. Another question. The information you find within the phone. But if the information transform to the cloud, what do you say about that?

Yes. What do I say about that? First of all, it's not we find the information, it's our customers. So, we don't check any information. And today, what you have is normally a mix between mobile data and cloud data. So, what we help our customers with is, for instance, to find the password to the cloud service. And we -- and that's normally in the phone. So, if we can do -- use our tool, the Ramanalyze, for instance, which checks the ROM memory of the phone, we can find the password to, for instance, WhatsApp in the cloud. And then we can give that to our -- or the customer can use that then to access the data in the cloud. But if everything is in the cloud and nothing is on the mobile phone, then we don't -- but that is not the case. I think it's going -- it will be a mix.

Today is a mix.

Yes. Today is a mix, and I think it will continue to be a mix.

All right. Thank you. I think we have to leave it there. Thank you, gentlemen. Let's take a short break, and we will be back in 20 minutes.

And also, if you have more questions, talk to Tony and me in the breaks or in the lunch or wherever, whenever. [Break][Presentation]

So welcome back, everyone. I hope you had a good break. We'll now continue the program with a session on market and customer presented by Marten Blixt. And with that, Marten, the floor is yours.

Hi, everyone. Really, really happy to meet all of you here today. My name is Marten Blixt, and I'm the Chief Commercial Officer here at MSAB. I'm actually quite new to the business. I started 3 months ago in beginning of February, but I'm super excited and happy to what I've been seeing and be part of this transformation and growth journey that we have ahead of us. And what I've seen during my brief tenure here is that we have a growing market, a proven product and a commercial organization that is really ready to scale. So, let's start looking at the market, how it looks today and what are the main growth drivers. First of all, it is a market with strong structural tailwinds that really plays to our advantage for MSAB. Some of the growth drivers that we're seeing is, of course, the data explosion. And I'm not just talking about number of volumes of devices. I'm talking about the pure volume within each device. We're not talking about gigabytes and megabytes any longer. We're talking about terabytes within each of the phones. Secondly, it's the encryption and complexity. It's getting harder and harder to actually extract the information from each of the devices. Thirdly, the geopolitical realignment. We're seeing an increased need from customers that are looking for trustworthy, European and with long-term experience in the field. And then last, the AI-enabled workflow that we're seeing entering our space as well. And as Peter and Tony are saying, the size of our company, our adaptability and flexibility makes it easier for us to incorporate and implement AI in many parts of our product and organizational workflow. And if we look at the competitive landscape, you saw this pie chart from Peter's presentation before, but I'd like to digest a little bit what it means to us and how it plays to our favor that we are the runner up in this game. First of all, it's a stable three vendor race. And as Tony talked about, the churn in the market is super low, and it's hard for competitors outside of the top 3 to enter this space. Secondly, it's clear that no vendor rules them all. So, you actually need all of the top 3 to solve these cases, and that highly plays to our advantage where we have now an extremely potent product and are able to grow and take market share. And what I've been seeing is also our fantastic complete offering from frontline to the digital lab. So, we have a really compelling competitive advantage where we cover the entire customers' workflow in multiple different customers' use cases. And last but not least, being in sales, the larger part of my life, what I've seen with our existing accounts, the number of accounts that we have to grow and penetrate within, sell more number of licenses, sell more products and also sell the complete end-to-end offering. That is one major growth lever we have that plays to our advantage. Secondly, the emerging markets. When I've been out meeting customers all over the world during this brief tenure, I've seen the need, and they're building up their forensics labs, their frontline offices for the first time, and they really need someone like MSAB to lean upon doing so. So why are customers choosing us? There are multiple reasons, but I try to dissect some of them. First of all, we're European and trusted. We're Swedish listed and European headquartered. And it plays to our advantage when it comes to being compliant with all European data regulations out there. Secondly, our frontline advantage. There is a clear trend and a theme where police officer and law enforcement wants to solve the crime closer to where the crime has been committed. The use case and how you use our product is a little bit different there. But we, thanks to the product development team, have a really potent product to address that specific need. And it's a clear differentiation compared to some of our competitors that we have this complete end-to-end workflow. Thirdly, our global footprint. As Peter is saying, we have customers in over 100 different countries, and we have local representation in all of the major geos, Americas, Europe and Asia Pacific. And I'll talk a little bit more about the products, but our complete product portfolio is extremely interesting and rewarding and add value to the customers. We're having the right product at the right time. I'll talk more about XRY Pro, XAMN and UNIFY in a slide or 2. So, who do we serve today? This is -- Peter talked about the various customer types, law enforcement, government, military and defense. And here, you have an understanding of the coverage. But looking at public safety. That is definitely our heritage, our DNA and a core in our offering. Civil government is expanding. Governmental agencies, central intelligence. It's an area of interest for us, which we believe is going to grow significantly in the years to come. Private sector is a sensitive topic, as Peter mentioned. We want to make sure that our product is being used for the right reasons, definitely. But there is also a clear white space for us. Our competitors are there, and we want a piece of that lunch. And there's a clear white space opportunities, and it doesn't really do any major changes from the product point of view. So that's an area we want to tap into definitely. And the beauty of this, if I was an investor, is that we don't depend on one single customer. So, we are spread quite broadly. We are not depending on one single product. And the product evolution, as Tomas will show you later on, is really, really compelling. And the feedback we're getting from customers are super positive. So, what are then the products we're selling? I just want to take you through some of it on the high level. We have our XR -- XRY product. To make it easier, it's the product we are using when we have a mobile phone that is unlocked. Secondly, the upgrade and the growth journey we've had over the last couple of quarters is our XRY Pro. Now we're talking about more advanced capabilities. We're talking about not only Android; we're talking about iOS. We're not just Samsung, but also Apple. It's when you need to enter locked phones. So, we're talking about password lengths that are severely long. We do brute forcing to make sure that we really, really can extract the full file system from these mobile phones. Secondly, we have examined. It's where you do the analytics parts. When you extract the data from a mobile phone and there are thousands and thousands of chat messages or pictures and you want to find this image with a gun or with a certain individual. Our tool and analytics to find this information is best of the best. And last but definitely not least, UNIFY, the product to be used to collaborate. And when I've been talking to some law enforcement today, they're using USB sticks. They're having hard drives, SD drives to send and share. It's highly vulnerability. UNIFY helps agencies across the world and across and within the agency to collaborate, share and reduce significant time and money as they solve these type of crimes. So, I'd like to take you through how a typical deal comes together. I thought that it could be an interesting way for you to understand and to see. So, starting with the law enforcement here. And to make it easier or simple to understand, it's pretty much twofold. One, we have the law enforcement and the lab community. And secondly, you have the frontline police. And of course, their use cases is vastly different. So here, we're doing complex case extractions in the lab to a very fast triage need in the frontline police officers. We're leading with XRY Pro, XAMN and UNIFY within the lab, but they're using our tools as XRY Pro Express or Kiosk in the Frontline. Looking at our customs and border agencies. As Peter mentioned, they have a very short period of time when they have and detain an individual in the border. They're talking about minutes that they need to extract information and take a quick decision, should we keep this individual or not. So, using our XRY Express and Kiosk solutions with built-in workflow to support the agent at hand to make a fast and right decision. Secondly, the military and defense units. We're talking about defense-grade triage solutions, that are military adjusted and adapted. We talked about rugged also hardware, but also, again, when you're being shot at, you need to make sure to take quick and easy and correct decisions. We have a similar use case as a law enforcement. So, some on the analytics side on the military, but definitely on the frontline police and the more offense part, as Peter is also mentioning. And last but not least, the enterprise space. Here, we talk about corporate security, insurance companies, banks, et cetera. They're mostly often are working with mobile phones that are unlocked. So, we lead with XRY, XRY Pro and some other parts of our offerings as well. So, the beauty in this is that we have both the breadth as well as the depth in each of our customer segments that we're focusing on. From sales, what we have been working on and will be working on is to make sure that we have a really structured way of going to market in each of these verticals and that we support sales as well as our partners ensuring that we push and lead the sales journey. So, we adapted our sales journey to the customer buying cycle. Who is our audience, what are they looking for and how are we selling to it? So, we're following it to identify, qualify, evaluate, procure and then deploy solutions. So, we have a structured way of adapting into each of these customer segments. So where does the growth sit? I have taken the 2025 revenue results from MSABs last year and try to dissect where what is driving growth in each of the geos and what is our focus within. So, starting with our foundation market, EMEA, representing 54% of the total revenue. We'll see a clear growth driver where they're favoring European first vendors. It's a perfect match to us. We're seeing increased budget coming from defense and border controls, and we're seeing a modernization going on and again, more into the frontline use cases, as I talked about. So, our focus will replicate the growth pattern we're seeing in Europe, and we play that to fit our strength in the business. So, government and defense border use cases. So, we're building up a repository on that to make sure that we attract, and we are easy to understand what we're doing and how we're solving each of these use cases. But more importantly, EMEA is our foundation market. Here is where we have the largest number of existing accounts, and we have a massive opportunity to grow within each of the existing accounts, upgrade to XRY Pro, move over to XAMN and then the full suite with UNIFY, and they need it and they need it badly. Secondly, Americas, our scaling market. So, what's driving growth there? Well, it's the largest digital forensics market in the world. We need to be there. We need to actively proceed with new customer wins there. There is a trend of growing federal investments in the market as well as within the Americas regions, we have LatAm as a geography that is untapped and super exciting to be part of. So, our focus, we need to be there when we are being evaluated in each of these cases. So, we have done account mapping with every single of our competitors. So, we know when the renewal is up, what is installed, what are the price mechanisms, and we want to be part of that evaluation process for sure. Secondly, the federal enterprise deployments. Here is a classic land and expand, but we know who is being funded. We want to be part of those conversations, and the team is really aligned and set up to be part of this growth trajectory onwards. And last, LatAm, of course, been part of the growth channel. And I can just say that we had 2 weeks ago, we had a partner event in Athens, where we had 70 partners from over 40 countries. And the feedback from those partners on a grade 0 to 5 and 5 was the best, the average score was 4.9. And that is a beauty for me to see that what we are showing, what we're talking about to the customers, how the product development that the investments has been doing over time is being perceived from these customers, more importantly, in this case, from the partner's prespective. Last but definitely not least, APAC, the emerging market. We are investors, you have seen the growth trajectory we are on to in that market, and we feel extremely excited based on the conversations we're having in the region. It's the fastest-growing region in the world as in regards to the forensics market. And many of these customers, they're building their forensics labs for the very first time. And they need someone trustworthy. They're looking for -- really love a European supplier. And as Peter was saying as well, Android plays to our strength in that area. Coming in as new, I like to challenge the Android piece at least. We -- that's a definite thing for us, but that won't be enough if we're going to reach the SEK 1 billion in 2030. We need to be as potent and as skilled in Android as iOS as well. And the conversation I've seen so far, it seems very promising. So, focusing our APAC, we have a direct sales force focusing on the more direct markets like Japan and Australia and the other markets we're working with our extended arm in our partner ecosystem. So, when I were looking through the entire global market, I've identified 3 engines of growth. And those are focusing on where we're strong. We have 1,000 customers in 100 countries. Three priorities. Number one, upgrade, get every single one of our XRY customers up to XRY Pro, and we know the value they will see from this upgrade. Secondly, more seats. Once they're in, once they're hooked, once they see the beauty of our technology and the results from these extractions, we're supporting them on, we're going to spread within the agencies. And thirdly, the development that we have done in XAMN and UNIFY moved from just one product to more products to the full platform suite. It's exciting times to speak to our existing customers and see their feedback to us. Second growth engine, new markets. And I have to say that I'm super impressed that in every single geo we're in, we're having emerging markets. We're having new markets, even in Europe. So, APAC, again, fastest-growing region. The response we're seeing extremely positive. Americas, Latin America, some of the largest deal we closed in Q1 as well as we're working on for Q2 will come from LatAm market, super interesting and nice to see. And EMEA, yes, we have some mature markets, the Nordics, U.K., the DACH region, France, to mention a few, the Netherlands as well. But there are other countries outside of Europe that are really, really interested in working with us, and we are definitely targeting those markets as well. And last but definitely not least, we have been talking about verticals as a highly important part of the puzzle. And I'd like to mention three verticals. First, the trend we're seeing with the frontline police. That's definitely something that we are different compared to our competitors. It's a huge political use case because the labs currently, they've been overflown and drowning with lots of devices. So, we can solve and reduce that workload by solving the crime closer to the crime is committed, we're making a huge impact, plus we're reducing efforts on the forensics digital labs. Secondly, the military and defense it's a great market for us. Our focus here will be to make sure that we are structured in the way we go to that market. There's a little difference in vocabulary and how we are penetrating those markets. But we have the product, we have the technology, and we have the resources of translating those messages. And the same really is for enterprise, definitely white space for us. We are definitely going after that market as well. So, three engines of growth leads up to some sort of a conclusion. Where do we go from here? Well, as a summary from my point of view, the market and the market drivers that I spoke to you about before really plays to our advantage, local supplier, trustworthy and our capabilities around frontline. Secondly, our customer base, thousands customers, spread over 100 countries. That's a massive opportunity for us to go deeper and wider within plus the opportunities we have in the emerging markets. And last and absolutely not least, 3 months in, the commercial transformation is prioritized. It is resourced. We know what to do. The market is moving, and we, for sure, are doing so as well. But don't take my word for it. The best thing we learn and evolve as an organization is by working close in relationship to our customers. And it's with great pleasure that I'm introducing Arnaud from the France law enforcement, who will take us through his views from a daily forensics' operations in one of the largest police forces in the world. So please put your hands together and make Mr. Arnaud Bristielle welcome. Coming all the way from France. We're so happy to have you here. Arnaud, welcome.

Thank you very much. So, thank you very much for inviting me. I will today try to share with you the use that I made of the MSAB solution with a small presentation. I will start by introducing myself and the main mission that I handle, make a quick overview of the digital forensics, okay? And then show you our MSAB enable mission with a case study. So, I'm a sergeant major in the French Gendarmerie. I work as digital forensics investigator. I have 23 years of experience in law enforcement, and I spent 16 years in digital forensics. My main mission in order of involvement, forensic analysis of digital devices such as computers, mobile phone, IoTs, USB storage, anything with digital data on it. Support to search operation. We go every week on field to help to our colleague to perform search operation to detect the device and identify them. We provide technical assistance to the investigation. When my colleague has difficulties to find evidences or when they don't know how they can find evidence in the cloud or in Internet before we are going to the arrest phase, we'll help them to gather information. And I will also perform technical monitoring and process adaptation. We are in a moving field. Every day, there is a new solution, new application, new hardware. So, you need to stay sharp and to always be looking for the new things. And I'll also train and transfer knowledge to the front investigator to help them to get the data on field. I work -- the main type of offenses that I handle are on the upper end of the spectrum. Most of them are rape and sexual assaults against minor, offenses related to child sexual abuse materials, so downloading video and photos with children on it, serious crime like homicide, very violent crime, things like that and also organized crime, drug trafficking and theft and things like that. Rape and offenses related to child sexual abuse are 80% to 85% of the case we handled. We are involved during all the investigation. During the initial investigation, we support field investigators, as I said, to the digital investigation looking for, I don't know, IP address, things like that, help them to interpret data provided by site like Facebook or things like that as they collect sometimes quite a lot of data. So, we have them to sort them and find what is important. During the arrest and search operation, we will go on field. That's real photo from case we are handling. We are there to identify devices, okay, because sometimes for a field investigator that seems not important or sometimes too old, we go there and say, no, we can't find evidence on that. We need to take it to the lab. We also put the devices in the right condition. If you need to -- especially with mobile devices, you need to perform some things on field to be able to extract the data later. So, we are there to do that. After the arrest of the guy, during the police custody phase, we will deploy labs in the units we are assisting to quickly uncover initial evidences so the judge can, in the first hour, take decision against the perpetrator. And during the follow-up investigation, when the case is a bit complex or we have some issues with the devices, we'll go to the lab, and we will sometimes disassemble a computer to get the hard drive or sometimes change part on mobile phone because on this case, the USB port was not working. Without USB port, we cannot extract data. So, we ask for spare devices, and we switch the component or in this case, I was able to recover the bullet shot by a guy, so we can try to extract the data from the phone later by replacing some part on it. A quick digital forensic overview. So, when is digital evidence important? Almost every crime leaves digital trace. In violent crime, we are looking for location, pictures and video. We have often people who are filming or taking pictures of aggression of violent crimes on the street. Chat exploitation, for chat exploitation, we are looking for, of course, media, but also Internet navigation, messaging and things like that. Cybercrime, we're looking for hacking tools, ransomware, tools for phishing and other stuff. For drug trafficking, communication, location and financial crime, you are looking for spreadsheets, mail and things and so on. Why is digital evidence is crucial in front of court. Those are facts, not opinion. It's based on data, not on human memory, okay? You can say what you want to do. It's very difficult to remember something precisely 1 or 2 years before. In phone, it's not a problem. The data will stay as the same. It's persistent because the data can remain for years, sometimes over decades. If, for example, you have rape videos stored on a old hard drive, the hard drive stays in a shelf or something like that. 10 years later, the data is always on. We just need to plug the hard drive in, and we can get the information. There is a massive data collection, especially since the introduction of smartphones. Your smartphone is for a user -- I don't remember the things -- excuse me, in French. No. To help you to get more from your phone, your data, we store a lot of information like location, the way you type on your keyboard, some terms and things like that. And there is a very large number of data collection. So, it's very interesting for us. And it's often invisible to user because in the mobile devices, you are not -- you don't have access to all the data of your phone. If you are using, for example, a Windows, you can go to the Windows folder and see what's inside. You cannot do that on an Android phone or an iOS phone. A lot of the data is invisible. And even if you delete things, for example, you can delete pictures you made from the crimes or things like that. You go on the bin, okay, it's completely deleted. You think that, but we are able to recover some thumbs or things like that in the system side. What challenge we faced in the mobile forensics? During the acquisition phase, we have more and more device locking and encryption. And it's so very difficult to get the data. We need tools for that. We have no administrator access. Even if I have the password of the phone, I'm like a user. I cannot go in the system files. I cannot go in the system folder, so I cannot find directly the hidden treasure, the hidden evidence, so I need a tool. And as I just said, it's -- most of these data are hidden or inaccessible. And during the analysis phase, we have a massive data volume. It's usual to get 2 million pictures on a phone extraction. That's pretty common, okay? So -- and that's the same for location, message and things like that. Many, many data are stored in a way that is not human readable. Before you can interpret the data, you need to process them to decode them so they are -- you can exploit -- not exploit, you can work on it. And the structure of the data is continuously evolving. You can see on your phone; you have regularly new version of your application. And so, you need to be -- to stay sharp and to be up to date. And it's difficult when you are in front end. That's the reason why we need tools. So, I will share with you a case study. It should have some difficult thing to hear. So, I prefer to say that before I start. So, this case study is very recent. I've handled it in the late March 2026. At first solicitation of a minor and during the investigation, we discovered sexual assaults involving a minor. So, the initial findings, the suspect is a school bus driver, okay? He's 30 years old, nothing -- no bad particular background. He is accused of sending inappropriate messages to a girl aged 12 to 14 that he transported. At first, he asked a number of the young girl to say, okay, I can tell you when I'm late or when I am in advance, so you don't need to wait in the rain. And then he say, "You were cute today." I want to hug you; do you like me? Maybe we can see each other outside and things like that. But the messages were deleted by the victims. And we were not able to recover those messages because the lack of space of disk space available on the device. As the victims' phones was -- have a very few storage, the data were erased and replaced by other ones. So, we decided to put -- to go on early morning arrest at the suspect's residence, 6:00 a.m. We cannot go earlier. And to use the surprise to prevent the evidence destruction. We always said at 6:00 a.m. when the police knock on your door, your brain juice is not fresh yet. So, you don't have all the things aligned. The goal is to place the guy in the police custody for interview. Seize all his devices and during the custody proceed to the examination to find some evidence. We do that even though we think it's easy in a sense, because if you think, you don't have a lot of evidence against this guy and you don't put all the -- everything on it, you don't take this seriously. You cannot say that he is innocent. Maybe there is -- there will be people who said, no, you are not innocent. The police didn't put all the effort to catch you. So, they lost you, okay? So, during the arrest phase, the suspect took 10 minutes to open the door. As we don't have a lot of evidence, we didn't break the door, okay? When we tell him why we are there, he showed no sign of surprise at all. In fact, he tells me, okay, I understand, but she got no evidence. That's really not the way some people tells you said that, okay, well, we'll see that. We found one device on him, okay? He gives us immediately the password, and we can continue to -- continue our investigation. We didn't know that he was a single father of a 5-year-old son. So, we get -- we handed this young boy to the family, so they can take care of him during the custody of the father. It's important for later. Okay. So, after the other thing, we are -- during the search phase, -- we found a second phone on the bedroom, power on, locked and charging on the bedside. The suspect appeared suddenly very uncomfortable, okay? And said, I don't have the code of the -- I don't use this phone anymore. I don't have the PIN code, say, that's quite strange. I don't have a cell phone plugged, powered on my beds side, and I don't remember the code, especially if it's my old phone, most of the people keep the same passcode on the phone, but that was not the case. It was not the same passcode say, okay, that's not a problem. We will handle that later. We will end the search operation. So, the first phone analysis, we perform a full data extraction using XRY Pro, including system data that's not accessible to the user. And I was able to say that he deleted a folder at 06:05. We knocked on door 06:00 and he opened at 06:10 just before he was arrested. Over 200 CSAM file in this deleted folder, videos, most of this were videos with children between 18 months to 8 years old, okay, having sex with adult that he get from Internet. And Snapchat conversation with the victims, for the reason why we are at first there. So, we recover on this phone, the deleted conversation. The most important -- the most interesting thing were on the second phone. On the second phone, we were able to unlock the devices using MSAB XRY Pro. Full data extraction performed as well, including system data and user data. And we found evidence that the father sexually abusing his son. More than 20 videos recorded on the devices. That's the reason why he didn't -- as I said before, his brain juice was a little bit fresh, but not fresh enough to think about the second phone. Maybe you think he thought that we will not search this one. In conclusion, an effective tactics and tools enabled to recover data -- deleted evidences related to the initial case, okay? We got the messages and messages over little girl and uncover intrafamilial sexual abuse. After that, we can safeguard the 5-year-old boy to his family. The father was directly in jail, waiting for his trial, all within 48 hours. So, in something very quick. So, imagine when you arrive and you know that the children were abused to be so effective in the term of timeline is very, very important. Thank you for your attention. I have nothing else to say.

Thank you, Mr. Bristielle, and thank you, Marten, also. So, we will now move on to the products and the underlying technology. And for that, I hand over to Tomas Taesler.

Thank you. I'm Chief Product Officer at MSAB. I joined the company about 1.5 years ago. And I'm going to give you a bit of a more introduction to the different products of our portfolio. And I'm going to start doing that by illustrating the use of products through a video that sort of is showing -- what could be a real case from extraction to sort of analysis and arresting a criminal. So, I take you back to the city of Grayport. Let's imagine that's in South of England in December last year. And the police station gets a pretty routine call about the car accident happening. So, they send the officers to the scene to investigate. And when they get there, they discover that the driver is missing and the passenger is unconscious and not cooperating. They soon realize that this car is actually stolen. So, they collect evidence from the car, two mobile devices, from the passenger which is turned on but locked and one from the supposed-to-be driver, who's missing but left this device into the car, which is then also locked and turned off. So, they bring these devices to the police office and do an examination. They connect the passenger device to XRY to try to figure out who this guy is, where they have been potentially connecting to where this car has been stolen, still thinking this is only a car theft. They can do selective extractions for the last 7 days to examine that content on his device. And when they have that content, they move that into XAMN, the analytics tool to look into who is the owner of the device, looking into all the content available, choosing to look for geolocations where this device has been moving around. They see some interesting connections with sites where there have been other crimes committed. And they look into the message history to find out there are some interesting message correspondence between the owner of this device and another person called Anota. In the message log, they can see the actual messages exchanged between the two people and something indicating that these guys has come over quite a lot of cash. So now this opens up to something else. It reveals a connection to an armed robbery, actually, that happened a couple of weeks ago. So that, of course, initiates an interest in the other device found also that was turned off. So, they bring the evidence back to the Grayport forensic lab, where they have XRY Pro, and the forensic experts can try to break into the other device. For that purpose, they use an application we call BruteStrom Surge. This is a GPU-powered hash cache-based brute forcing application. And after a while, they're able to crack the passcode of this device to get into analyze all the data available on that one. So, with cracked password, the locked device can be accessed, and they can now start to connect the information from the two devices, see how they have been communicating, with whom they've been communicating and what they have been communicating about. And as they do so, they actually find one of the messages an address that looks interesting. It happened to be an interaction between Mr. Justan and Anota and a message confirming a delivery address for food delivery. They take that as an indication of a site of interest, they get the search warrant, and they go to this address and find evidence from the robbery. So together with Arnaud's presentation and this illustration, I think that gives you a good picture of how our products are used and the purpose of the business we're in. The portfolio we provide is addressing all the steps in the workflow of a police department, law enforcement organization from ceasing information on the field to bringing devices to the lab for more advanced analysis to the investigator office for trying to connect the dots and find what is very relevance to creating a report that holds up in court. Important through this change, especially for law enforcement, that what is presented in court can be proven was unchanged from when it was ceased on the device, of course. So, there are ways in this process to guarantee that nothing is manipulated along the way. And in addition to that, we then also have, of course, management tools for network management to help the IT department to operate the software on the endpoints that are installed with any law enforcement organization. Basically, the use cases applied here is the same also for defense and special ops type of customers, just that the requirements on the, sort of, evidence is not as hard as it is in law enforcement, of course. So, looking at the different steps of this value chain, as mentioned, at the core of it is getting the data. And with the data, we mean everything you have on the phone. We can do this by different means. We can politely ask the operating system on the phone to give us the data they want to share. As I said, that's pretty limited on mobile devices. It's not like the PC where you have full access to everything on the workstation. So instead, we use both physical extractions, and we use full file system extractions by getting underneath the hood of the operating system. The media we capture is everything from geographical information, as you saw in this example, where had this device been located? How has this person been moving around? We can track messaging through a wide range of messaging apps, of course, call data records, media of all types, pictures, movies, whatever you have on the phone, passwords used for logging into networks. You will be surprised when you see the extraction that we can present on XAMN, the vast amount of information available on the phone. All your life is on that phone, and everything is accessible for us basically. So, when we do this extraction, we put the information into a secure container that we call the .xry file. That is the entity that goes on in the investigation going forward. That is the format, proprietary format that is then locked down and secure and audited for any type of reading that is done on it. So, we can prove that nothing was changed from extraction to the courtroom. So, this is what XRY does. And we package that for different purposes for use in the field, we package XRY into what we call XRY Tablet or XRY Express, which is sort of a laptop-based solution for the purpose of being able to capture data already at the crime scene or for that matter, also to get data easily from victims and witnesses. Imagine you have been witness of a crime. You maybe have a picture you want to share, but you don't want to hand off your phone for a couple of weeks to be sent to a lab. You don't want the police to have access to all the information on your phone. You just want them to have that picture. We can support that by selective extractions on site at the victims or the or the witness location. The other option is the Kiosk that I think was mentioned. The XRY Kiosk is sort of more stationary frontline equipment, which offers you a little bit more capable interfaces for data extraction that typically then is deployed at the police station, which was used in this case in the movie where the device was brought into police station and they could do the examination there. All of this gives you that folder, the same folder then as being extracted in the lab, where we can do more advanced extractions. Typically, the device is locked. You want to get into a specific type of device and operating system that is not as easily done as with some other devices. So, this is where the forensic experts are working. They have more competence and then more powerful tools also to work with. We also use -- here, we use XRY Pro, and we also use that when we help customers through our access services where our own consultants can help them to get in data from phones that maybe they have tried and failed with themselves. Quite often, we can do tweaks to the solutions that is released for customers to get into even more devices than what is otherwise available. The simple sort of positioning of XRY and XRY Pro is that XRY gives access to unlocked Android. It's a little bit simplified. We do actually get into locked Android as well, but it depends on what type of device it is and so on. Unlocked iOS with sort of a limited extraction. That's kind of the polite asking the system to provide this information. Feature phones, burner phones, more often used in crimes than maybe anyone in this room would think. So quite an important sort of range of devices to have support for and different accessories. We talked about drones, we talk memory cards, other type of associated accessors that hold information that we can get into also. XRY Pro does everything what XRY does, plus gets into locked Androids, complete extraction, full file system extraction of unlocked iPhones and also ROM and memory extraction. So that's a simple sort of differentiation between XRY and XRY Pro. Once we have that container of extracted data, we move it to the investigator's office and start looking for the evidence, connecting the dots, finding that picture that holds the evidence they needed in the stack. And for that, we use XAMN Pro. It's a very capable analytics tool that gives you tailored filtering and search possibilities for mobile forensic purposes. It helps you to follow conversations between different parties, as you could see in the movie here also, independent of what different messaging apps they have been communicating across. You can see the complete chain of communication for two or more parties and follow sort of the conversation that they may have had. You can see how people connect to each other. You can see everyone that I have been communicating with and everyone that those I have communicated with have communicated with, so you can follow the chains of people relationships. Geolocations, as I mentioned, and yes, everything available on the device is then possible to search and filter and analyze through XAMN Pro. If you want to do that investigation at a larger scale, you may have 50, 100 gigabytes of data on a device, even more. There's thousands of pictures, as Arnaud mentioned. It's just too much data for one person to dig into. You have a team of investigators working on different focus areas in parallel. We're offering the UNIFY Suite as a way of collaboration to make it possible for more investigators to address the same data at the same time and create one consolidated report that can go to the court for presentation of evidence. Of course, this is an area where AI becomes very powerful. It's a lot of data to be analyzed. So already today and on our road map, we have a lot of initiatives where we are utilizing AI to make this XAMN Pro more powerful in terms of being able to find what is relevant for the investigation. We, of course, have media content recognition to find specific type of media automatically by content recognition, pictures, including CSAM material, guns, drugs, face recognition, those kind of things. Of course, also language models for transcription and translation, language detection. More and more now looking with the more powerful large language models to have natural language queries to have even more powerful searching and filtering of information and also fake media detection tools. Also in research, I mean, what makes it possible for us to get the data depends on our research. This is dependent on knowing how the hardware and the operating system on any device is working together for us to find the vulnerabilities on that, that can help us get into device even if it's locked. And also, if we get into device, actually being able to access the full file system extraction. So, we use AI more and more now in our research to sort of elevate the performance of our very skilled researchers. And then, of course, as in any software business, agentic development for application development, we see that with AI, we can scale the number of flavors of applications, use cases we can support for specific customer segments and use cases. Then in the court dimension, XAMN produces a report that holds as evidence for a court case. And on top of that, we have the network management tools to help the IT department to operate an installed base at the customer site. This is still very much an on-prem business. We're starting now to have more requests coming for the cloud deployments. We do support cloud, but our customers for legal reasons very often are very conservative in terms of storing this kind of data in the cloud environment. So typically, they have the need to operate this installed network themselves. And this can range from just a few nodes with XRY and XAMN licenses installed to several hundred at a big customer. So of course, you need efficient ways to update the software when we do releases, to manage user access, to update watch lists and hash list on those clients that are doing extractions and everything can then be done centrally in an efficient way. Looking forward, some of the key investment themes we are seeing is, of course, always continued the research and development of our access capabilities in XRY. Tools for quick triage, as mentioned, as one of our growth avenues also, field triage solutions for specific use cases military special operations, not the least, where you quickly can seize a device, see if there's something of interest on that device, maybe looking for some specific contact names, phone numbers, pictures, having the tools that gives you a quick way of on the field getting like a green light or red light that this is of interest or not. Cloud deployment, going into more of an optimization of the platform and the whole solution for that. Client network scalability to maybe to handle even bigger states of installed applications and also then, of course, in XAMN AI augmented analysis tools. So that's, in a nutshell, what was in our product portfolio. Thank you very much.

Thank you, Tomas. Please stay on stage. We'll now move into a panel discussion. Maybe we can have some help to get the chairs. Okay. And for that, I'd like to invite CTO, Anders Jonson. Please join us. And we are also joined by Jim Dowling, CEO of Hopsworks, an AI expert and lecturer at KTH Royal Institute of Technology. And Jim is working with companies like Saab and Ericsson on developing large language models or more specific LLM solutions. So welcome, everyone. let's spend a few minutes on AI and what it means for MSAB and the broader space. Let me start broadly. How is AI changing the industry, Jim?

Digital forensics industry. I mean what we saw earlier in some of the demos were analytics. I think, first and foremost, I had a quick look at the demo out there. And it's great you can search around and find things, but I think natural language has become an interface that people can use to dive deep quickly, find correlations, find patterns with natural language. So, I think it's going to improve the productivity of the people using these tools, such as MSAB. I think that's the main kind of thing we're going to see here in the near future. In the long term, agents and LLMs are going to change many industries.

Do you agree, Anders?

I do. From the customer side, absolutely. I mean, I think with a growing volume of data in the software or in the phones, there will be a different way of accessing the data. And there need to be new workflows, which is aided by AI definitely. But then since I'm the CTO, I think the way of working is also changing inside when we produce the software. And in R&D, we have basically two departments. It's research and research is using AI even more than they did just a couple of months ago actually. And development, as Tomas mentioned, we are using agentic engineering to sort of leverage the play field compared to the competitors, so we can produce more with the amount of stuff we actually have today, which is pretty cool.

Do you want to add something, Tomas?

Yes, the product perspective in that case it's obvious with the kind of data volumes we're handling and the different types of data that AI can simply be a superpower for investigators.

But is it like everything happens faster than you expected?

I mean I can take this. I mean I've been around a long time. I mean I've seen the Internet bubble. I've seen the mobile phone revolution. I've seen deep learning come and I've been working in AI for 30 years. But this is beyond anything. The change even in the last 6 months. So, my company, Hopsworks, we've gone from some people playing around with coding agents 6 months ago to 95%, 99% of our code is written with coding agents now. And I guess we're at 5x the productivity of before. So, software is going to be built incredibly quickly and with high quality, not just -- it's not going to be just slop. I mean a lot of people think it's slop, but you can actually engineer this thing. So, we're going to see a lot of changes in how software is built and deployed. That's going to be the main thing. And it's going to affect a lot of workflows that may not have been digital before, they'll be digitalized. And I'm sure that's going to affect digital forensics. I mean if we take the example of digital forensics, we have large data volumes. And typically, how we process large data volumes is we need to search through that data, and we might have something like an index, and there's many different types of indexes from database technologies. LLMs won't change that, but they'll change the way you interface, and you access that data, primarily through natural language. So, if you have data analysts that need to learn how to use SQL and build dashboards, that will be revolutionized as well as the interfaces to all digital forensics' products in the near future.

How do the rest of you see it, Anders?

Can you repeat the question?

Yes, if it happened faster than you expected.

Yes, most definitely. I mean a while back; we had 4 releases per year. Now we are down to basically biweekly. This is in response to the increased release cadence from the different vendors we are sort of competing against. So, we are releasing more often. And also, I mean, the sort of, it's become a bit more problematic. I mean, Peter mentioned in his open speech that you heard about Mythos, I guess, and Anthropic's new model. It has increased the difficulty of sort of answering the more frequent releases from the vendors. But we, in turn, have started to use AI internally more to -- in response to that. So, we are releasing more frequently. And yes.

But is it like hard to -- for the companies to adapt to the path?

It could be, I guess. So, it has to do with the amount of investment you can do, but also, of course, you need to be very, very, very skilled. So, we got 20 researchers in our department, a total of 19 in my department in total. But those 20 persons are very, very skilled. We're actually recruiting globally. So, they're not living in Stockholm, all of them, in order to find the best ones in order to have -- in order to respond to the increased difficulty basically.

And Tomas, will you be able to adapt the technology?

I mean it's a trick question because the speed of development in terms of increased capacity for large language models the last year or two or whatever timeframe you look at this is basically logarithmic. So, it's hard for the human brain to adapt to logarithmic development, right? It's just -- it's a matter of catching up rather than foresee what will happen. But clearly, what we -- the capabilities that AI offers today compared to just a year ago is mind-blowing when it comes to how that can be applied in a mobile forensic solution for analysis and research and so on.

Yes, exactly. So, in 5 years, where will we be?

You tell me.

All right. So, I just also have to ask you, maybe start with Jim about Mythos. Yes. I mean, is it a threat or an opportunity or ...?

I think it's the near-term threat to a lot of system software companies, including my own and MSAB. I think there's two parts to it. One is that large language models now work with any task that is what we call verifiable. So, if I have a program and I can write it, run it and verify that it works as expected or it could be mathematics, any tasks that we have out there that are verifiable and could be based on natural language will be nailed by LLMs. So that's kind of the first part. And programming has been the first main one that we've been talking about. Mythos has taken it to an incredible extent. So, for us, internally, we're like scrambling to kind of do threat analysis of our software because when some of these are released, the bad actors will, of course -- and the bad actors can be your competitors so that can basically -- you can down a competitor with this. So, it's not necessarily going to be just -- I think if we take back to mobile forensics, when models become so much better, it's also the users it will affect. So, they're going to spin out messaging apps with crypto that you haven't seen before, and they're going to be even harder to crack. And then we have the quantum issue as well, quantum of computing. So, there's a lot of -- I think it's an industry with a lot of kind of change will happen in the next few years.

What do the others -- what do you think about?

I mean it's -- it's definitely a kind of threat. It is absolutely. But I mean, MSAB has been doing this for 20 years. So, playing this sort of attack-defense game, we are quite skilled at it. We are not so nervous. But yes, it's a new landscape, absolutely. But again, we are prepared for it.

But will it make your moat stronger or weaker?

We have had a lot of discussions internally regarding this and the reply from research is actually mostly positive. They say, yes, it's more difficult, but we are using the same tools as our competitors or -- and the vendors. And in some instances, we can increase our effectiveness tenfold or even more depending on what area we are doing it. But it's also produces a lot of new possibilities.

Do you have any thoughts about Mythos, Tomas?

Yes, who doesn't. I think...

Should we be afraid?

No. This is a great opportunity, to be honest, it gives us as a small company, as Anders says, the same powerful tools to do research as anyone else in the industry. So, we are levering the odds a little bit with that, in my opinion. And also, when you look at what we do and for the purpose we're doing it, what any large language model can do today technically may for different parties in this ecosystem, not be interesting or profitable. The business case looks different for us versus the vendors versus our competitors. And where we have a good business case to use the full power of large language models for research and product development. The vendors of mobile devices may not always have that. So, it remains to be seen sort of how that power game will sort of end up. But as Anders said, this is -- that is not really a dramatic change. The sort of the power balance has been there for years.

But where do you actually see AI creating the most value in your products?

If I may answer that, clearly, in analysis, for all the reasons I described. It just gives us the investigators a superpower in terms of finding the needle in haystack that otherwise you would have to have many investigators in the police offices to try and do. You have -- I mean, the biggest iPhones holds 1 terabyte of data, I think. It's just too much to -- for one person to go through. With AI analytics tools, you can all of a sudden change that completely. So, in law enforcement, you'll need less people to do the same analysis. In research, it just gives us a superpower to do more. And in product application development with agentic development -- sorry, vibe coding or agentic engineering or whatever it's called now, which name changes monthly. We -- I see an opportunity we can do more flavors of the applications we have -- the moat is to extract the data. That's very hard for anyone to copy. And it's both hardware and software architecture combined. So, I don't really see that being threatened by AI. The capability of getting the data is our moat. And on top of that, we can now all of a sudden build much more flavors of applications for specific user segments that basically just helps expanding our offering.

Anders, what's your take on that? Or do you agree?

I totally agree with Tomas. And I think, I mean, looking at the competitors, Cellebrite and Magnet foremost, I mean, yes, they are bigger, but with the help of agentic engineering, I mean -- and the amount of token we are willing to use, we can basically produce the same output as they do now. So that's only a positive thing.

But just to clarify, what's the biggest risk with AI actually? Jim?

Risks. I mean there's risks in -- if you're in the area of digital forensics, there's risks that there's going to be new attack vectors that are going to appear that will overwhelm the ability of law enforcement to actually deal with them, right? So, everyone is using WhatsApp, Facebook, Signal and so on. And I guess you've cracked all of them, maybe not Signal, I don't know about that. But the new software will be just put out there that will have very good quality encryption and there won't be existing tools there to crack it. So, it's -- I guess the biggest risk for the industry as a whole is that new tools will appear at a rate that you can't keep up with. I do think because of the agentic engineering, you will be able to move a lot faster and keep up. But the arms race will get faster. I think that's the biggest probably most likely outcome.

All right. Your industry, MSAB industry, law enforcement and the military tends to be quite skeptical in some markets like U.K. How do you break through? Tomas, maybe.

In what way are the skeptical?

No, but I mean, there have been several cases where evidence has dismissed, because it wasn't evaluated by...

I know. Typically, I would say it's often in the hands of how you have used sort of public AI tools in sort of -- yes, in a not so clever way sometimes that you have used it to -- which has led to -- in the U.K., there's been some situations where they have went to court with evidence of what produced by AI and we just don't do that in our products. We use AI to help find what could be of interest. We never manipulate; we never draw conclusions. We just expose it for the investigator to do the judgment. So rightly used, it is not an issue from a sort of any point of view. But there will be some cases which has influenced the public debate in the U.K., for instance, that has been a bit unfortunate.

Okay. And there's a lot of -- I mean, we've touched on this, but there's a lot of talk about AI replacing software and SaaS companies. And could AI replace parts of what MSAB does today, Anders?

I would say no. The reason is that we are in a very specific market, requires knowledge and skill, especially to extract the data that requires humans to create the exploits. I know Mythos, Anthropic, says they created exploits themselves. But the ones we are sort of finding are very, very difficult to find. So that won't be replaced by AI anytime soon. Then also to create the softwares we are creating, we need to be very, very careful about automating the code generation. We need to review every line of code in order to have a software which is guaranteed to produce court admissible stuff basically with holds up in court. And if we automate the entire production line by AI, that would be pretty unfortunate, I think. So that makes this sort of business area, it creates a bigger moat, and we are not that replaceable, which is a good thing for MSAB.

Is it like that, Jim?

I can't comment specifically. I only know the tool briefly. I can't comment about SaaS in general because, I mean, at our company, we do marketing with a tool called HubSpot and pay them EUR 3,000 a month. Our marketing person, who is not a programmer, built an equivalent tool and sort of integrated a couple of open-source tools. It wasn't too much work, a few weeks' work, pretty profitable. We've replaced a few other SaaS applications like a managed webpage, Webflow, something ourselves. Now our marketing people are writing the website with cloud code. Because it's easier to -- if you're making changes, it's spread across the whole website and so on, it's very easy to do those globally and it would have been much harder on the SaaS platform. So, I think a lot of SaaS software is under threat. But I don't know how fast it's going to happen. It will happen, but it will be industry kind of related, right? I think in different industries, like it will go faster and others, it will go slower.

So -- but where is the real boundary if there is any?

I don't know if there is a boundary to be honest. I mean we're a database company. We build a database, right? This is the lowest of low engine. Our code is open source, RonDB is a database. And we -- our Head of database, Mikael Ronstrom, he's written 150,000 lines of code in the last 2 months. He's written a query optimizer that he hadn't touched. He's been working at different layers, and it's really good. And this is the hardest code you would expect to write with agentic coding, but it's actually quite good. So, I don't know where the layer is. I really don't know where the moat is.

Do you want to add something, Tomas?

Yes. No, every software that is exposed on the Internet user interface-heavy will be copied. Can be copied, will be copied. So, the question is what you have beneath the hood that is not exposed and copyable. That is what the value you can protect and that's what you can build on. I mean I get contacted regularly by analytics companies who want access to our data. We're not giving them our data. So simple as that. So, what applications we build and expose -- yes, I agree. I don't see why they could not be copied. But the question is they will just be empty shelves if you don't have the data.

Are you agreeing?

I do.

Okay. You don't want to add anything. Okay. But as talking about competition as a smaller player, how do you stay competitive in an AI-driven market? I mean, there will be many competitors.

Should I answer? I don't think there will be actually. We have three on this market. And the barrier of entry is increasing -- getting quick -- more difficult quickly, which is to our advantage. Smaller companies using or creating exploits single-handedly, et cetera, they are -- is my guess going to disappear. So, I think MSAB is in a very, very good position to scale and to compete using AI also, of course.

So, Jim, do you agree that this favors specialists?

I mean I think definitely, if you have a niche that has a high barrier to entry and you have a good moat, that's not going to be the first software company be appended. I mean there's tons of workflows that are not digitalized today that will be digitalized in the very near future. There's tons of incumbents that basically charge huge, huge margins for something that you can vibe engineer. Those will be appended first. So, it's going to be a journey. And I mean, I saw, for example, Legora, who are a Stockholm company who are massive. And I saw someone the other day, well, I've open sourced the vibe coded version of Legora. Now how true is that? I don't know. But I think there will be a deflationary aspect to AI and software. If you're -- we have a competitor called Databricks, they're valued at $150 billion. They charge huge margins, and we can do the same thing. We could be the Lego version of data AI analytics. And when times are good, maybe it's fine for databases. But when times don't turn a bit and you need to save cash, there will be a deflation aspect to this, I think.

Right. So good to know for the investors. In other words, you're saying AI can replace others, but not you. All right. What is one thing investors might underestimate about AI in this space. Anders, do you want to start?

Underestimate. That's a difficult -- not prepared for that one. I'm drawing a blank here.

Do you want to...

Well, I'm underestimating AI every day. So, I guess investors are too.

Jim, do you want to help us?

I just think that there will be more attack vectors and maybe going to need more forensic and digital forensics in the future. It's going to be a higher cost if it's going to be easier for people to hide digital trails.

I mean with Jim here also, I mean, there's a saying there is no such thing as a perfect code. And I do think that applies to AI, too. We will see more apps popping up in both Android and iOS and more apps is more attack surface. So that may be to our advantage actually.

But what is like the weirdest thing you think we will use AI for that we like can't imagine right now?

Well, I have been sort of theorizing about maybe the police will be using agents themselves. So maybe humans aren't clicking through our software, who knows in the future because the amount of data. So maybe they just point an agent to our software and say, find evidence for this case, whatever and present this when you're done. Voila, it sort of presents this, hopefully holding in court as well. Maybe that's a future, but not, but that would be interesting.

Jim, what do you think?

I work in the area of data, right, scalable data. And in data, we typically say data is either structured. So, it has -- it's stored in tables and it's organized and you can search through it. And then we have this big bucket we call unstructured data, and that's everything from files to text to images to videos and so on. I think AI is going to organize that unstructured data in the next few years. I think in theory the idea that we're going to have these like big dumps of unstructured data that no one has a handle on and there's a lot of tooling around. I think we're going to just organize all that.

So that's the weirdest thing you think we're going to use.

It's weird for me, it's weird. If you work in data, it's like it's a revolution, but yes.

But do you think, Tomas?

The most weird thing about AI.

No, no. But I mean the most rare thing that we are going to use AI that we can't really imagine.

Yes. Well, that's -- Personally, it's already -- I see people using AI as a psychologist and so on. That's to me crazy. Within forensics, I don't know. It's not really -- I don't see any crazy use cases. But again, I don't have seen it all yet.

I wrote a book on AI for O'Reilly, just came out in December. And I kind of described LLMs as it's an amnesiac brain and a jar. I'll try and break that down. So, it's a brain in a jar in the sense that large language models are brain. They understand language, and they have language going in and language going out, but they're not connected to the physical world. They're not connected to. They don't have vision. They don't have sensors that we have. They're also amnesiac because I can tell something has forgotten it immediately. It's like that movie Memento, write it down. But he had better memory than LLM because he wrote stuff on his body. to, but they don't. So, at the moment, they're amnesiac brains in a jar. And I think when they get connected to the physical world and they're interfaced into things; we're going to see big changes. Yes.

Interesting. Maybe we have some AI questions in the audience.

[Foreign Language]

[Foreign Language]

The -- yes, the professor at KTH, Hannes Alfven, he wrote this book about -- actually, it was the computer writing the book about the development of the computer, the big computer. And at that time, of course, the humans, they were so dangerous that they were kept in lakes or very, very, very carefully taken care of. So actually, the humans were only the way to produce the great computer. This book was written in about the '60s by the Nobel Prize winner KTH Professor Hannes Alfven, but you didn't know about it. When will it come true?

The question was... when will it come true?

I mean there are many people think that that's the next step in evolution that we're just a steppingstone. A lot of people in the tech field aren't particularly unhappy about it. They just think it's a natural evolution. But that's not my opinion, but that's what I hear.

Okay. Do we have any more questions? No. Okay. So, guys, is there anything else you want to add on AI that you think it's good to take with us from today?

I can ask if you're not programming, I expect you will be programming in the next 12 to 24 months. I think most people in most jobs will be producing software of some kind with -- powered by large language models. So rather than wait for the future to come to you, maybe you want to move to that future earlier. That would be my take.

Tomas, do you have any closing remarks? No?

Just say that internally at MSAB, we are quite energized with the dawn of agentic engineering. And I know some of our product owners in the audience here, they're also using agentic Engineering and sort of have started using and producing proof of concepts, real software just to show that something actually is working, which they then can hand over to R&D for proper development. But it creates a quick cycle of new things actually, which is it's very energizing and everyone can be active sort of in the development, which is very fun.

Great.

That actually makes me really tired because I'm not sleeping because I'm so much programming. That's another side effect.

Okay. Thank you, everyone. That brings us to the end of today's session. And before I hand over to Peter for some closing remarks, I just wanted to tell you that the panel will be here for a quick chat, and there will be light lunch served and some products you're welcome to have a look at. And with that, I hand over to Peter to wrap things up. Thank you very much.

Thank you. Yes. AI, both a threat and an opportunity, and we are keeping our eyes on both. I can promise you that. But talk to these guys during lunch. When we had dinner with Arnaud yesterday, he told us -- first, he told us about he lives in this beautiful little village where they locally provide very good things that you can eat and drink. It seems to be a fantastic place. But he also told us that he brings a sleeping bag to work. A sleeping bag to work? Yes, because he has only 48 hours to find the evidence for the criminal. So, it's not the time that you go home and sleep. That's the kind of commitment our customers have and the crimes that he solves pretty important. So that is what we -- the customers we serve with our products. And I think that is a reason to invest in MSAB. But maybe you want more reasons. So, I mean, I put up five reasons. We have a good strategy that we are executing on, and it seems to be working. We have a good organization in place with a lot of new people, new energy, but not only a new organization, we also have changed the way the company operates. We -- today, we have the data we need to take important decisions. We have more efficient processes and so forth. We have great products. We have really, really top-of-the-line products today. We don't have a lot of products, but the ones we have, they are really great products. We have the growth avenues that we see in front of us that we talked to you about, both in geographically, in the products and new verticals. We haven't stepped into those yet, and we are still growing with really, really good speed. So, I think there's a lot of reasons to invest in MSAB. You have invested this half day in MSAB, and I want to thank you for that. And I hope you got some value for it. So, thank you for coming. Now go to the lunch, spend it talking to the people, the management people we have here. They are all really easy to talk to you. Thank you.