Yes, hello. Good morning, everybody. My name is Jukka Tapaninen, Aiforia CEO, and welcome, everybody, to Aiforia's 2023 earnings announcement. Today, I will cover the results in brief, and then we go a little bit deeper on the financial review and then cover the strategy and business targets going forward. At the end, we have a session for Q&A. 2023, it was actually really interesting for the company. We actually get a lot of progress in many areas. As many of you know, one of our biggest kind of referenceable customers in the U.S. is Mayo Clinic. What happened last year is really important that we got live on a clinical product on a breast cancer side. Since that, they've been kind of taking in use a couple of additional models. That was really kind of a groundbreaking event for us. So that happened at the end of Q1 last year. And now it's in real-life production. And as you know, in Mayo Clinic, there's over 2.6 million slides a year. So now we are, of course, covering just a kind of a small piece of that volume. But that's actually kind of ramping up very quickly. The other big event last year was the pharma sector overall. We made a couple of nice deals. One was the Orion in Finland. And then there was another one in Switzerland, a big one. And in the pharma sector, they can actually leverage a lot of AI going through the kind of huge amount of samples that they need to analyze for the drug development processes. But that's a really important sector for us. And that was a really kind of nice progress that we did last year. So then on the regulation side, we started the FDA process. And we are planning to get some results in 2024. So that's really important. So, of course, in the U.S., we can already sell to the clinical side using the LDT, the laboratory-developed test process that some of the hospitals are using. So it doesn't block our sales. But we think that when the market moves more on a volume phase, then the FDA approval is really important. And, of course, we got the HIPAA compliance last year. So that means that we can host the patient data in our servers in the U.S. So that was also kind of a nice achievement from last year. And then since the IPO, what we did in December 2021, we set up some short-term targets for two years. And at the end of last year, the two-year period was coming up. And so we actually nicely achieved... And I covered those later. But we achieved all those results that we set up during the IPO. And that was actually referring to the title of this presentation so that we are doing kind of steady progress and so far moving forward as planned. So some of the deals. Yes, again, the Mayo. And as I said, in the breast cancer side, we are live in the clinical side. But what is also important is that in the research organization, there's over 70 pathologists using our kind of tools to develop own AI models. And there's over 30 ongoing research projects going on. And we see this year the outcome of these things. So there will be some new AI models in collaboration with the Mayo. So there will be a lot of scientific publications coming out on that side as well. So the relationship with Mayo is strengthening. One good example was the launch of the QuantCRC model for colorectal cancer. That's a model that is developed by Mayo. We have exclusive rights to commercialize that. That's probably the next level of AI in pathology. First in the world. But it's a predictive model. Combining the data and diagnosis that we can do with the AI. We are collecting 15 different data points. Mayo has built a predictive model on top of that and forecasts the patient outcome. This is going to the next step. Because with the other portfolio what we have, the kind of key argument is that we can make the process more efficient and less mistakes. We can automate. But with the predictive model, we go closer to the treatment decision. Based on this information that they're getting from QuantCRC, they make decisions on how much chemotherapy the patient is getting. And, of course, they can downgrade the amount. And it's really beneficial for the patient to get the right amount because it's a hard treatment. But on the other hand, it's financially interesting as well because just in the US, there are 150,000 cases per year. One month of chemo costs $10,000 a month. And if you can lower that amount... They kind of calculated that system savings, if everybody would be using this, would be something like $900 million just from one AI model. Because they calculated out that 30% of the patients could be actually getting less chemo than previously decided. But that's where the kind of AI is heading. And this is a solution. It's actually getting a lot of kind of attention in the community. We hope to build a nice business out of it. Realistically, the first customers are happening in Q2 or Q3 this year. Then we start ramping up the business after that. Of course, in the US, we have multiple other customers as well. In Europe last year, we closed the biggest deal so far, EUR 1.2 million. It's the Veneto region. Those Finnish listeners, you can maybe kind of think about the numbers, what I'm telling. In the Veneto region, there are 5.5 million people, 12 hospitals, three centers for pathology. In the first phase, they started analyzing 200,000 cases. So that's the first phase of the contract. They bought all the five CE-IVD-marked AI models what we have. So now it's in the implementation phase. So we have delivered the first piece. And with that customer, the volume ramps up at the second half of this year because they are now doing kind of acquisition for the scanners. So they have now some scanners that they could implement the system and start using our solutions. But later on this year, they will ramp up the volume because they do a big scanner acquisition. And in Italy, there are +10 similar size kind of hospital regions. So there are some kind of sales cycles going on. Catania, we closed. So that was a nice one. And there will be more to come on this area. But I think the point here is that we have a really nice referenceable customer in the U.S. I'm sorry, in the U.S. and in Europe as well. And then the PathLAKE. So we won the first deals there. So we were kind of gaining traction on lung cancer and prostate cancer. On lung cancer, we are the only one that NHS Trust can buy at the moment. And now they have a second tender going on for additional AI models. So now we are covering... If we think about our key markets, the U.S., Europe, and U.K., we have good clinical customers on each one of those. And then, of course, we have some opportunistic customers outside of our key markets. But these are really important achievements from last year. And in results, Veli-Matti, our CFO, will cover those more in detail. But I would summarize that we grew roughly about 50% last year on revenues. We did the change from a Finnish accounting system to IFRS, but still grew 50% on the order book. So remaining performance obligations that the customers have towards us. So that was growing over 100%. So this is kind of revenue that we are kind of... All kind of money and deals that we are kind of making as revenue in the next coming years. But these are contractually already covered. On the cost side, we stayed more or less in the same level as in the previous year. And that was the plan. When we did the IPO in 2022, we started to kind of build up the organization. And some of the recruitment was happening in the second half. So now in the first time, comparing the second half of 2022 and 2023, we have the similar cost levels. So we have now enough people to execute what we need to execute. And that's about it, the first piece. And after the financial review, I will cover the strategy and ongoing activities. Veli-Matti, please. Thanks. Good morning to everyone. My name is Veli-Matti Parkkonen, I'm the CFO of Aiforia. This is the first year that we are reporting according to IFRS. The biggest surprise for me was that the report grew to almost 80 pages long. But seriously, there were changes in our numbers due to the transition to IFRS. We announced those changes in a company release a couple of weeks ago. But I will go through the main points here one more time. Revenue recognition. Actually, we made the decision to recognize revenue according to IFRS in the beginning of last year. So there weren't that big changes throughout the year. But all in all, the transition postponed the revenue recognition compared to FAS. Share-based payments, that burdens our P&L. On the balance sheet, that is offset with a similar-sized entry to equity. In the balance sheet, the effect is zero, but burdening on the P&L. This is non-monetary employee costs. It doesn't affect the cash flow. Leases. When you make a lease contract or rental contract more than 12 months or one year, you have to book that as a debt or liability to balance sheet. On the other side of the balance sheet, you book that as an asset, right-to-use asset. The balance sheet grows slightly. On the P&L side, we used to book rental costs as normal operating costs. Now we are booking those as interest payments for the lease debt and depreciation of the right-to-use asset. On the P&L side, the effect is almost zero. There are some other changes also, but they are not affecting our figures that much. I won't go through in this presentation those changes. Figures. Jukka mentioned some of these figures. The main points here. Revenue grew 49%. It bases mainly on the license revenue. Stable growth, add-on type of growth. We don't have that much project type of... More volatile project type of sales. On the EBITDA side, it was pretty flat year on year and showing good improvements towards the end of last year. That reflected also in EBIT figures, although the bigger depreciations made the effect a little bit smaller. Equity ratio was 66%. Cash. We had EUR 14 million at the end of last year. That plus the EUR 3 million underwritten Business Finland loan facility, EUR 17 million altogether, gives us a good base going forward. However, we are looking at opportunities to strengthen further our financial base. Order book, as Jukka mentioned, grew more than 100% to EUR 2.4 million. Last year, we had on average 73 persons. Personal expenses pretty much reflected that growth in the amount of persons working for us. Investments were almost 40% of our cost base, mainly in R&D. We pay a lot of attention to the costs. While the employee costs still were on the growth phase, the other operating expenses decreased even more, what the employee costs were, towards the end of the year. We have made some adjustments at the end of last year so that we are aiming below EUR 15 million fixed cost level for this year. That compared to EUR 16.3 million last year. Somebody might ask that you are burning too much cash. Well, there is a really fine balance between the company health and the competitive edge. There is a great market opportunity, and you have to stay ahead of the competition on that. So I could assure that the company management is on top of that challenge. Then comes the cash flow on operations, which we promised to be break-even by the end of 2025. Pretty good progress throughout the year, being in the first part -EUR 4.7 million and improving to -EUR 3.5 million at the latter part of the year. This paves quite a good way to the other promise to be positive the year 2027. This also shows a good indication that we are a pretty asset-light business, and we don't need any working capital for our growth. And finally, the board of directors' proposal for the distribution of profits. So the board of directors proposes the general meeting on 4th of April, that the loss for the financial year would be retained in the retained earnings account, and there won't be any dividends paid. Back to you, Jukka. Thank you, Veli-Matti. So the market opportunity. I've been saying this a couple of times, but we are in a market where there are customers who have an active problem recognition, and they are actively looking for solutions. Meaning that, okay, even though we bring a new innovation to our way of doing things, we are not inventing the market or market need. The market need is there. So they are just kind of actively looking for solutions to sort out their issues. The cancer rates are going up. The number of pathologists actually stays the same, decreasing compared to the rest of the physicians. And the workload is going up because as the cancer rates are going up and those cases are studied more carefully, there are more slides to be diagnosed. So the demand is there. And the cost is there. And the market is growing. So the digitization is happening. So there are 3 billion, depending a little bit from the source, 2.5-3 billion slides that need to be digitized. And then those digitized slides need to be analyzed and taken diagnosed in the traditional way, looking from the screen or looking with the microscope. But eventually, the volume is too big, so you need to automate, and AI is the solution for that. So the demand is there. And the breakthrough is happening as we speak. So there's a lot of public tenders from European countries coming up. And big institutions are picking up this topic. And we also decided that, okay, so for us, it makes sense to win the big ones. So we are kind of focusing also on the top 10 hospitals in the world. I would say that plus five are using us. A couple of those are already kind of good customers of us. But they are the ones who are leading the change. I think the next one will be that when the solutions are a little bit more packaged together, then the market will start scaling up fairly quickly. Our position on this workflow is something that... Okay, so in the case of a kind of a tissue sample, there's always a physical sample taken. They put some staining, prepare the samples on a glass plate, put some staining on it. Then it will be scanned. After the scanning starts, the software piece. There's an image management system, workflow systems. Finally comes to a diagnosis. And we believe that the diagnosis is the area where the value is created because everything in this process actually aims to that point so that, okay, you actually understand what's on those slides. And we are the core of that business. But also we kind of work together with all the other vendors. We have the interoperability. We have different strategic alliances and agreements with the multiple vendors so that we can offer kind of a full solution, end-to-end solution to the customers. From our side, thinking about what is our leadership. So we were actually the first one to start building a deep learning-based AI solution for pathology back in 2015. So we came up with the first tools in 2018. So as you know, we started to build a platform so that we enable pathologists to build their own AI models. That's one of the core assets of the company still today. In 2020, we started to build our own AI models to kind of go more on a clinical market, which is financially probably a little bit more interesting than just the preclinical sector. That's the killer combination so that we have the tools to build the models, and we have also the ready-made models. That gives us an edge because we are not offering a black box. The pathologists know where the results are coming. What we do... we actually show the results. We do all the calculations. The pathologists, they have an opportunity to check the results, even change the results, and there will be an audit trail. We are top of the technology leadership. Now we are looking up, and we are already leveraging some of the Gen AI things on our development. But as Veli-Matti said, some of our investment really goes on keeping that edge. So we are recognized on the technology side as the strongest player in the market. Scalability. So of course, this is an interesting business model because once you do the implementation like Mayo or Veneto or some other hospital... So when it's up and running, there's a kind of cloud implementation, all the security implementations, all the kind of connections to surrounding systems. Once it's there, so I think it's a really sticky process. And then you can start adding volume with the different AI models. And then you can scale it up. Interoperability. So we are agnostic towards the laboratory information systems or image management system or scanners. We work with all of those. We've been testing and building these kind of solutions together with the other players. The last point on this one is the certification and the quality and the security. It sounds like a small thing, but in healthcare, it's not. So it's actually kind of heavy lifting, what we've been doing. Getting the patents done, getting different certifications done on security and other types of things. So this is something that we really focus on going forward. As I mentioned in the beginning, the FDA and HIPAA compliance are such good examples of these types of things. Our strategy going forward. So I'm not going to cover all the details on this slide. So of course, we have those megatrends kind of supporting us so that, okay, things are really happening. We have an extremely good position to be the leader on this market. That's helping us. On the other hand, we need to be focused on certain things. We need to make sure that our product offering is top-notch. We have something that, okay, it's useful in the clinics, and it's useful in pharma. We listen carefully to our customers. We build and develop the products further. That's definitely on our focus going forward. Of course, growth. We are still fairly small numbers. Even though the revenue backlog is growing and so on, we are closing nice deals. We are really now heavily focused on growth and strengthening a little bit our own sales activities because we are realistic in that sense so that the first deals, we need to do it ourselves. Then we have an opportunity with the ecosystem to grow. That's why. So we are really kind of talking with all the leading players on the scanner side and others on this area so that we get the leverage on their connections and their solutions as well. But the growth is clearly the focus on the next couple of years. And people. So we have a good team globally, plus 70 people. So we are making sure that, okay, we have the top talent in the organization and invest on that side as well. Well, this is a history. So these are the things that we promised when we did the IPO. And as you can see, gaining the 5 customers, then large pharmaceuticals reach over 5,000 users. As you know, we have a presence in 50 different countries, plus 5,000 users. The only one that is a 5 out of 6 is the number of CE-IVD-marked solutions, AI models. We have actually a CE-IVD mark viewer, so we could have said so that we have a six out of six. There was a change in European regulation, so now it's IVDR. And there was 12 months that there was no notified body who could do this kind of certification. So now that problem has been sorted out. And we started to kind of do the first steps on the new process successfully, kind of passed those first steps. And you can expect that, okay, going forward, we will kind of bring out new certified models fairly quickly. These are the targets that we set up for the next coming two years. So of course, as I said, AI models. So we will do those. We will kind of bring those in the market. We certified. But we will focus also on a research use-only portfolio. We have actually already over 500 AI models done with our tools. We can find some additional research use models to kind of bring into the market. But also, we count on our partnership like Mayo. Mayo's research team is building their AI models, and we will start commercializing those as well, in a similar manner as we have done for the QuantCRC. There will be more AI models coming to the market. We just launched a product, Aiforia Studies. It's a study-centric workflow tool for good laboratory practices. That's a little bit kind of... It's not a core AI thing, but it's really needed in a preclinical setting. It's a kind of an audit trail type of a thing so that you have a workflow so you can audit all the changes and different things that happen on that process. We built that on customer demand. Now starting to sell the first deals this year. Clinical offering, strategic partnerships. We, of course, already have a partnership with Google. We have a partnership with Amazon and Microsoft because those are the three biggest cloud providers. Our solutions are running on the cloud. So that's going on. So whatever kind of a cloud setup customer wants to have, we can deliver through these partnerships. But then, of course, we have a kind of interoperability with Hologic, multiple others, and then the scanner manufacturers. This is an area so that we are going to focus more on the next coming two years. We believe that, okay, that's also kind of a key to kind of really access on a volume market. I mentioned the big accounts. We want to make sure that we have traction with the big hospital chains and hospitals in key markets and from a revenue perspective. It's a great opportunity to have kind of these customers, but also from a reference point of view. If you start from the top, it's easier to kind of scale up in the mid-market as well. Then Veli-Matti mentioned about the cash flow from operating activities. We are on track achieving all of these things. The previous targets, we achieved, and now we set up the new ones. That hopefully leads to the midterm targets by the end of the decade. This market, as I said, is opening up fast. AI is a hot topic. We see the traction from public tenders and customer discussions from the US that this is the hottest topic in pathology at the moment. EUR 100 million, yes, seems like a kind of a big target from current revenues. But actually, it's a realistic target. So there's 5-6 years to kind of get into that level and keep in mind so that we are in a SaaS business so that we start accumulating those customer relationships and annual recurring revenues so that in the year of 2030, we don't need to sell new EUR 100 million business on that year. So we have already kind of a backlog, which is helping us on achieving that goal. And then where it heads, like the QuantCRC model, so that, okay, now we are looking at the things that, okay, beyond the human eye because there are things that, okay, the human eye just cannot diagnose from those samples, but the AI will start picking up those types of topics and building up those predictions. And suddenly, then the value, what the AI model plus the predictive model are bringing into the game, it's much bigger than the efficiency and accuracy type of value creation, which is big as well. But okay, so then if we talk about QuantCRC, the price point for QuantCRC diagnosis is 10-20 times 20x comparing to the traditional AI model. The volume might be smaller, but then the case per pricing is much higher than in the standard AI model. And that's it. So we have an AGM coming on April 4th. You will find the presentations and all the information on our website. I think we have now time for some of the questions. Hej, Mikael Rautanen här från Inderes. Starting from the notified bodies and how the CE-IVDR pipeline is sort of going forward, you mentioned there's movement now happening. Should we expect that there's kind of a big block of models going through, or is it more of a chain of individual models that you might expect to go through over a period of time? How does it look like right now? Well, so I'm not going to give exact numbers. We have some thinking about the topic. Now the process... Because it used to be so that you kind of certify one model at a time. There was this CE-IVD process. In the new process, they actually kind of first certify the company. So checking that we have all the processes in place. They actually review so that we actually work according to those processes. And once that is done and we have done the first steps, as I said, so the next thing is that, okay, we can put multiple AI models in one certification round. So you can expect that, okay, in the next, let's say, six to eight months, we will come out with the multiple models from that. And then going forward, it's actually easier to bring those out. All right. But sort of waves maybe is the best guess at this point. Yeah, yeah, exactly. Yeah, that's helpful. You mentioned that in Europe, it's looking pretty good with the market activity. Obviously, with the CE-IVDR models coming out, there's more to sell to the market as well. But if we think of the US, of course, there's not maybe an individual FDA model in the market overall. Do you see that there is a sort of blockage for the broader market adoption in the US before we get FDA-approved models? Or do you think the kind of laboratory-developed test model that Mayo is using could already start achieving results in the US before the FDA pipeline is moving? Well, we can work with the LDT. And it's not blocking. But I would say that, okay, if we have an FDA, then probably the mass adoption of the AI would be easier. But it's not kind of blocking ourselves. And now we see traction from academic medical centers going forward. And then our offering, so combining this development platform plus the AI model, is a killer on that market because they want to have both. And they know what the AI is doing, so how it's trained. So it's supervised training. So they want to know what is the data that it's trained from. So in that sense, the current situation helps us on getting the first customers. And as I said, we have the FDA process started. And we should get some traction on that later on this year. But you cannot predict that exactly when it's going to happen because it's depending on the FDA. But so far, so good. Yeah. So I guess you're able to open customers in the US on the preclinical side, and that's sort of a buildup for later clinical use. But right now, the clinical use is maybe still taking a bit of time. Well, so Mayo is a good example so that it's not stopping us. So I think it's more kind of making sure that, okay, you have the right models. So they have a comfort that, okay, the models are working. So there's no problem with Mayo to adapt or some other customers, what we have, adding more models. Yeah, that's helpful. But sorry, one more point so that I think not the FDA, but so that we have a good kind of package offering to offer. So I think that's what is the main thing that, okay, when we scale up... But like in the Mayo, we have the Google Cloud implementation. There is a Sectra IMS system in place and Aiforia AI so that you need to package that offering so that it's easier to sell and easier to buy. So that then makes sense that, okay, it's easy to kind of put it in volume as well. Yeah. Continuing on Mayo, of course, I think now in the Mayo's official public catalog, there are three tests from Aiforia that are now live. And obviously, you mentioned that there's a big group of pathologists using Aiforia already. Is it fair to expect that there's quite a few models in the making behind the scenes at Mayo? And there's just a certain time that it takes for all these pathologists to get these projects live? How do you look at the ramp-up? Yeah, it's a good point because they started before the clinical side went live. There have been already kind of some publications coming out from that work. The idea is that now they are figuring out how they can get those AI models in production on a clinical side. Also, one way to do it is that, okay, we actually kind of certify those models or not even maybe use it as a research use only. But having a portfolio of Mayo-built AI models that we can start commercializing, which is then kind of funding some of the activities inside Mayo. But you can assume that there will be more coming out from that pipe. Right. Also on the predictive model side, hopefully. Absolutely. Yeah, that's good. That's good. Continuing on the US, I guess 1.5 years ago, you mentioned that you had City of Hope and Wake Forest coming in as customers. You haven't commented too much on these customers before or after that. But is it still fair to assume that there's some preclinical use in those customers? There is a Wake Forest, yes, definitely preclinical use. So we sell the Create. And that was the first step. So they were replacing our nice competitor from Denmark with the Aiforia. So that process takes a little bit of time. But okay, now Aiforia is in, and they kind of decommissioned the other solution there. But there's, of course, a huge growth opportunity. But now Wake Forest is with the Create, with City of Hope. So they have big plans. And so we will be moving forward with them. So they are already using the Create, of course. But so they will be kind of ramping up the clinical side as well. Right. All right. I think I'll let people on the chat line take some questions. Thank you. All right. Thank you. Yes. Let's take a couple of questions from the chat. First one, how has the collaboration with Mayo Clinic developed? Well, so from early days, we had a couple of leading pathologists started to use the Create platform. And as I said, it's something that enables the pathologists to build the models. And they like the solution a lot. That was probably the reason that we won the big deal 2.5 years ago because there were already internal users that they know the solution. They like the solution. And that was the starting point. Because in that tender, actually, we beat all the kind of competitive solutions. And that was the starting point. And then having that preclinical and clinical side covered, that was the key. So it's moving forward so that it helps the kind of scientists there to kind of build the scientific publications and build the models faster and nicer. So I think it's kind of getting deeper and deeper. And last year, we had a big kickoff event in Mayo. So there were about 300 people somehow involved in the Aiforia case. So it's some of them using us and some of them in other roles. But there's a lot of kind of investment from the Mayo side, the money that they pay for us, but the money that they actually have been investing on internal resources on leveraging the tools as well. Thank you. And the next question is about the financial outlook. Even though you don't give any guidance for this year, could you please elaborate a little bit on the outlook for 2024? Well, as you said, we don't give any outlooks going forward yet because one thing is, of course, that the size of the top line is quite small. So there could be a lot of volatility. And of course, on the other side, we could manipulate or affect on the cost side quite well. So we could handle one part of the outlook but not the other one that much. Of course, we have projections, and we have some ideas what the top line is going to be. And we know. But yet, I mean, we don't want, at this stage, to give any kind of precise outlooks. I have given out the idea what the cost side would be. So you could have that as a guidance. But on the top line, it's something that we don't want, at this stage, to give out. Maybe the best guidance is that at the end of 2025, we are operatively cash flow positive. So that's what we are planning to do. And that means that if we kind of stay with a similar kind of cost level, then we need to ramp up the other part, the sales. But it's a little bit... I wouldn't say binary, but okay, because we are dealing with the bigger customers. So it's kind of... We suddenly go up nicely, and it may be flat. But okay, the key thing is really so that we gain some new interesting customers because those Mayos and the new ones in Europe, they really kind of give us an opportunity to sell EUR millions for those customers on an annual basis. And that's the ramp-up what we need to do. But we need to win those kind of bigger organizations. Thank you. Those were all the questions from the line. Okay. Very good. So thanks, everybody.