All right. Hi, everyone. I'm Joe Verwink. I cover vertical software at Baird. Our next presentation comes from Certara, the leader in biosimulation software and services helping to revolutionize drug discovery and development. We have William Feehery, CEO, John Gallagher, CFO, with us on stage. This is going to be a fireside chat. To begin, I'll turn it over to Bill for an intro to Certara. Thanks for joining. Thanks a lot, Joe. It's great to be here. I'll just introduce Certara for anyone who hasn't come across us before, very briefly. We went public in 2020. We've reached about 1,500 employees right now. We operate globally everywhere the pharmaceutical industry does, is kind of the way you can think about this. One of the things we're very proud of is that 90% of all new drugs that have been approved in the last 10 years, we can trace directly to the use of our services or software based on public information around their filings in their drug applications. We're a profitable company. We're on 30% in EBITDA margin. We're roughly becoming a 50% software, 50% in services. You can see here some of the other statistics we have. Basically, the idea behind Certara is that we're dedicated to the proposition that the drug development process needs a new model. The reality is that nearly 90% of new medicines that enter clinical trials fail. Now, consider that's not just a random drug that enters into clinical trials. That's a drug that's been selected using all of modern technology, AI, screening, genetic information, you know, with the best idea going forward. They still fail quite considerably. That causes really the cost problem in pharma. Every drug that goes forward into any degree of clinical trials attracts a lot of cost. Most cost in drug development is actually in clinical trials. Some drugs go all the way to large phase III trials. They still have a 30% failure rate at that point, historically. If you look over the last, let's say, generation, 25 years, I'd love to tell you that it's been getting better as modern science has been getting better. In fact, it hasn't been. In fact, you'd argue it's been steady or maybe even declining a bit in terms of efficiency. You know, the drug industry is under some pressure now around costs and around that it's coming from price controls and things like that. It can be done differently. We have a biosimulation platform that consists of both services and software. Our platform spans basically the development phase of pharmaceuticals, starting in discovery, going all the way through approval. In discovery, we focus on the problem of which drug is likely to succeed as it gets into a big clinical population, which vary in, you know, all kinds of characteristics: age, weight, sex, comorbidities, genetic background. As we get into preclinical, we focus a lot on, you know, on dosing, what's the proper dose for this drug that is most likely to bring it forward and succeed in the largest population. As we get into clinical, we advise a lot on developing the right clinical trials to, you know, basically efficiently prove that the drug works. As you move into approval, we advise on how to get the drug approved at the FDA. Our drug development models are used by the FDA. Part of our software has been used by the FDA since the early 2000s. They're a big proponent of biosimulation. I'll just leave one chart, which is kind of the cool, you know, I've told you a little bit about the company and what we do. Here's a little bit of the cool science we do. One of the things we're really known for is our Simcyp Discovery Simulator. This is a biosimulation simulator we've developed over the last 25 years. This is quite a complex model of the human body. You can see it has all of the organs of the human body. What it does is it predicts the absorption, distribution, metabolism, and excretion of a drug, where it goes in the body, what the concentrations are. It's used for all kinds of questions as you move through drug development. This model has been used in, when it's used well, it eliminates the need for certain clinical trials. It saves our clients a lot of money. To date, we counted over 125 approved drugs where, if you look in the label claims, you can see the use of the simulator to avoid clinical trials and make label claims. There's nothing really else like that on the market in terms of the adoption and the use in the industry. We're still just getting started. There's a lot more we can do in the science here. We're investing still very heavily in this. I'll stop there and turn it over to Joe. Yeah, that's great. Any questions from the audience? Session two at RW Baird. I'll get those on the iPad. I'm going to ask a question. I've probably asked it every year you've attended the conference. To tee it up a little bit, we follow other technical application software where maybe the spending is 10% to 20% of that customer's R&D. Biosimulation is a very small piece of R&D in your industry. The TAM you define is only a point and a half or so of global biopharma R&D. Why isn't that a bigger number? Why isn't biosimulation spending a much bigger number? Yeah, good question. I'll answer it a couple of different ways. One is, you know, the pharmaceutical industry is a little bit different than some of the other industries. The pharmaceutical industry has been based on statistical science from its outset. There was a view in the beginning that human biology was simply too complex for modeling to make real recommendations. To some extent, we've had to battle that from our inception. The potential for the use of biosimulation as it exists today with no improvements is much greater than we've penetrated. Because we've proven this works in a lot of situations, it's also attracting a lot of investment. The technology is also improving. If you look at, you know, usually people tell me that we've overdefined the TAM. It's interesting you're saying we've underdefined it. You know, 1.5% of global pharmaceutical spending is an enormous amount of money. If you look at our product offerings, we've got software offerings in the discovery area. Companies are spending hundreds of millions of dollars a year on that. We are virtually the only real biosimulation solution in the clinical area. Obviously, that's half of R&D spending. Maybe we've underput. Our kind of view is we're a small part of a really big number that's out there. There's a lot more for us to go get. You mentioned the battle to educate around biosimulation. You know, month of April, something comes along, FDA, a bunch of exciting news that potentially eases the battle, let's say. Can we maybe talk about what new approach methods mean for Certara and specifically the aspects of your portfolio you think will grow much faster as a result? I think what you're referring to is the FDA announcement around eliminating animal models in, and they specifically mentioned monoclonal antibodies. They want to extend it past that if it can be done. We view that as kind of a recognition to some extent of what the reality is in terms of the science and also to some extent a goal or an ambition that's been set out there for the industry. Where we are today is we do a lot, you know, modeling, in fact, is used pretty extensively in monoclonal antibodies for a bunch of reasons. One is the models are pretty good. The other one is because, you know, giving a human antibody to a non-human primate and expecting to get really good data is perhaps a little bit far-fetched, right? I think the fact that the FDA has recognized that the science is getting to that level, and in many cases, perhaps not all, you could use that to reduce the number of animal testing. I think that's all positive for, it's positive, you know, ethically for the world. It's good for Certara. We're doing lots and lots of work. What we find is with our largest companies, most of them are doing modeling in parallel with animal testing. They would welcome a path to reduce the amount of animal testing because they're really basing a lot of their internal decisions on the modeling. The FDA has made that announcement. There are a lot of details that need to follow from that from a regulatory perspective. A lot of customers now are saying, OK, they've made the announcement. Exactly how do you do this? What do they want to see? What's the process? The FDA is still rolling it out, let's say. That's kind of where they are. A little bit more work needs to be done. There still has not been a customer who got a drug approved by the FDA under this new announcement. I assume the FDA will have some pilot programs and encourage some. That's still to come. This is maybe not the right analysis to be doing. You talked about QSP, which I think has the most immediate ramifications in terms of what customers would look to employ. Half of the bookings on your QSP solutions were deployed into monoclonal antibodies. I think about monoclonal as a share of all activity, it's maybe 15% to 20% of things that are in development right now. If that's 50% of your bookings, it just seems like any work happening there is basically using some sort of software. The share of work employing software has gone dramatically up. Is that the case? Is that kind of what you're seeing at the front line? Monoclonal antibodies is one area where the modeling has been particularly developed. The capabilities of the models are pretty good and widely recognized. It has become, I don't know if you call it standard, but over the last year or so, we've seen a big uptick there. We can do a lot of interesting things there. One of the questions we're asked on monoclonal antibodies has to do with first-in-human dosing. When you go into a human trial, you may have, let's say you're going to have 10 data points on your phase I clinical trial. You're going to modify the dose from some small number and move it up. If that range doesn't include the active range, the safe and active range, you've got a problem. If you've got five data points too low and five data points too high and only one in the middle, you have a problem. If you have all the doses too low, you're probably never going to go higher than the highest dose given. That's a really important question. You can answer that with QSP. That's one of the reasons why we're being brought in. If you get it wrong, you kill your drug. If you get it right, you can save a lot of money later with faster and better trials, for example. Yeah. This is going to create a lot of new product opportunities. Specifically on your QSP platform, you're doing some things. You've alluded to some new, it's really going to be a QSP platform. Can you just talk about what Certara is doing there? It's not going to just be on monoclonal antibodies anymore, but how that'll broaden out what QSP is used for? We acquired a company called Applied Biomath two years ago. That enabled Certara to create basically the largest group of people that do QSP modeling. QSP modeling, it's quantitative systems pharmacology. Maybe not the best term for this, but that's the most advanced modeling that's out there. We wanted to create a large group that were doing that, not only so that we had the capability of sort of serving customers, we had the breadth and the experience across lots of different areas, but also because we see an opportunity here to create an industry-standard software platform to do this. Right now, it's kind of the Wild West. People are writing their code in R, Python, C, everything you can think of. These are complicated models, right? You might be talking about thousands of differential equations. They're very hard for humans to understand. When you add on top of it, they're all written in, you know, everything else. It's a problem for companies. If someone leaves, all you've got is a SharePoint drive full of whatever the guy did. It's a problem for regulators because they're getting submissions in with every kind of documentation you can think of. It's a problem for us because we need a sort of defined process to create these models. It's not just creating the models that you get paid for. It's creating all the documentation around them. Where did you get this data? How did you backtest it? What's in here? What did you not consider because you think it was wrong? Where do you believe this model is accurate? All of that documentation around the model becomes a piece of it. Long story, but we're launching a piece of software called Certara IQ this fall. We believe that is going to become an industry standard around QSP. Why is it going to be an industry standard? Number one, we have the largest group of people doing this stuff, and they're all going to use it. We're going to have a good base of people both using and developing it. What the software does is it allows us to do the modeling. We've incorporated AI in it. We're basically leveraging AI with our experts to develop these models. We've incorporated all of the reporting and documentation that you need, what we call regulatory grade, so you can go to the FDA. You can say, look, here's all of the files that you would need to understand what's in this to understand whether you should allow it from a regulatory standpoint, which gives our customers a reason to come over here. We are very excited about it. We think that, you know, we have a great business in QSP. Right now, it's a services business. Like everything in Certara, we always look at things around services and software. We really are at heart a software company. We're creating the core biosimulation software, and then we array experts around that for companies that either, you know, don't have the internal expertise to do it or they don't have the very specific expertise you need to really make the most use of it. We're going to do that in this area as well. You talked on the second quarter call where you think the news out of the FDA and how that could evolve and create different decisions in the industry is maybe an incremental low single-digit billion, billions of potential new TAM for you. Would you say that QSP is specifically the biggest contributor of that increment? There are a couple of things. QSP is not just targeted towards reduction of animal models. It tends to be that that is one way to do it. We calculated that the industry spending on non-human primates is something like $5 billion a year. You could argue that's the TAM. Probably when you move from that to software, it'll be some number less. We've kind of picked something there, right? It's a significant opportunity for us to go forward. It's not just the reduction in cost. For a lot of our large customers, those animal trials take months. For a large drug, if you can take off a month or two in terms of the approval time, that adds up in terms of basically the NPV of the drug. That's probably where the real value is. Yeah. You mentioned a moment ago, Certara is sometimes at its best when you take the software, the AI/data infrastructure, but then you wrap your expert services around it. Services, of course, have had a lot of fits and starts over recent history, just given the macro and spending at your customers. It seems to be on a better path. The bookings recently have been very strong. How much would you maybe attribute just the end market becoming a bit more stable and that's conducive to you being able to execute better versus that business has gone through some different organizational changes? Is that the ultimate contributor? Yeah, I'll take that one. We don't, to be clear, we don't see it as the end market improving. Specifically, in the tier III category, we've been seeing strong performance throughout the year in services. We attribute that largely to the full build-out of the commercial team. We've been investing in sales and marketing. We had said at the end of last year, we thought we'd got that built out. We're seeing our ability to influence our results through that commercial team executing in an overall funding environment that has gotten a little better, but not a lot better. We don't see the end markets as the key driver there. When we look at tier Is, we had a good quarter in Q2 in the tier I category. As we look at the summer months, a little bit slower in tier I in the summer months, but we also see that as typical seasonality. We have full confidence in the plan for the remainder of the year. Maybe real quick, while we're talking about services, just the regulatory piece of the service business and where that stands in terms of finding a home outside of Certara? Yeah, the regulatory business has been performing well this year. We saw a return to growth in both bookings and in revenue, which is good. We have had a process going. That process has taken a little bit longer than we thought. We attribute that largely to some of the end market environment and some slowness there. Overall, we don't necessarily see the business as strategic as it used to be for Certara. We're continuing that process. We'll provide an update when we have a little bit more information. OK. I wanted to go back. Bill, you said something at the beginning about just how the TAM definition kind of gets debated. I think part of the debate is that there is spending on biosimulation. It might not go to an external third party. Customers are working on this to varying degrees. I'm just wondering, everything we've talked about to this point, whether it's different segments of your customer base, like you brought up strength in tier III, which seems more conducive to using the full turnkey offering, even just the importance of getting models into a drug development framework maybe more than has been the case, are those things all good for increasing your share of wallets? Would you imagine that more spending will end up coming your way? Do some of these events motivate competition and so that the share doesn't really move in a dramatic sense? Maybe it actually encourages others to do more than they have been doing in your categories? In terms of competition, the market kind of started out as primarily a bunch of small service providers. They're still out there. You can still hire people other than Certara to do this type of thing. The bulk of the people that really do this are in particularly the large pharma clients. Those companies have recognized the need for professional standard software that's not only professionally supported and supported with a lot of scientific background, but also is accepted by the regulators. We've kind of become that over time. It's hard to recreate that just because the way this works is the industry regulators want to know what's in the models. It takes a very long time to have them approved, right? Every year, we go through a process of expanding the software, explaining to the regulators what's in there, what wasn't in there, how good is it, where do we get the data to backtest it, and then finding a customer to bring forward those new features. That's been going on for 25 years. There's a lot of catching up, I think, to jump in. It's not to say we don't have, you know, it's a good market. We want to have some competition. We have a pretty big moat around what we do just because we have the kind of the first-mover advantage in terms of getting, you know, the industry-standard software that's out there. There's a reason for the industry to consolidate around something that people know and understand, in particular that the regulators are known to accept, right? That draws a lot of business to us. I wanted to shift gears and talk about AI. We were chatting on the way in that scientific software is maybe a good place to be in this debate because in just multi-industry enterprise software, the debate is AI is going to ruin the software business model. AI will make it easy to build software. In that debate, you have a view on the application layer and a view on the infrastructure layer. What I find so interesting about Certara is that you actually have both. You've built businesses and product sets in both the application layer and you have the data infrastructure. Maybe just explain the thinking there and how you intend to monetize. Do you have a view on which one could end up being more consequential? Yeah, so we're kind of on both sides of this, right? On the one side, we are fully leveraging AI. We've launched fully AI products, and we've launched features within quite a number of our products that add AI capabilities to it. The way we think about this is the core of what Certara does. We're basically building these very complicated differential equation-based models of the chemical kinetics of drugs in the human body and across populations of people that vary in a lot of ways. AI is not really capable of doing that, and even if it was, because we're set up in a way where the FDA requires very detailed understanding of what's in that, it would be very hard to just kind of turn loose a black box kind of AI model to do that. On the other side, if you look at how we do what we do, AI has really helped us a lot. We hire very rare high-end experts. You have to comb through lots of literature. It takes a lot of time. Their productivity has just gone way up because they can use AI to comb through all the scientific literature looking for that, looking for relationships between things, looking for specific data we need, explaining, doing some of the work around just sort of organizing all of that data. That's helping a ton. We've launched an AI product around reporting. Our scientists love that. Scientists, as far as I can tell, nobody really likes to write scientific reports or reports to explain the project to a client. There's a lot of productivity we've gained. On the scientific side, what we've started to do is take into account things that we can't easily model. There are some things where we don't really understand the underlying mechanisms, but you know they matter. An example would be real-world data. You have tons of data on what really happens when drugs are given across a population. We can use AI to add those into our models. Genetic data, we certainly understand that certain genetic patterns cause diseases, but we don't always know the underlying mechanisms. We can add those in. It's also added an ability to extend what we do. Regarding your data infrastructure, you have a platform there that I think is kind of being positioned as the data platform for your customers to do their own AI work and to be very extensible across their organization. Just the business case there. In terms of using, sorry. The Layer platform. Oh, Layer platform. Yeah, yeah. We've acquired an AI platform that lets you basically tap into lots of different data sources and organize them for AI. That's become the basis of us pulling together all of our software across the full span of the full development timeline into one platform. What we can do with that software is, effectively, you can index databases for AI without pulling all the data out of the database. When we bought that, we thought it was pretty cool as AI, and we bought it right before ChatGPT made all their announcement. As that has kind of rolled forward, what we've seen is, you know, companies have realized that they want to employ AI across the pharmaceutical area. If you look at some of our large clients, they have databases of clinical trial data that date back 50 years. They wouldn't like to apply it. They don't want to let that out of the company. A lot of it's poorly organized. Using Layer to kind of create an AI, the underlying data infrastructure platform to use AI has become an important part of that too. Great. Maybe just to finish on a few financial topics, the growth at one point in time was very consistently kind of in that 10% to 15% bands. Recent years have been below that band. What are some of the incremental things that need to happen to get you back into kind of the double-digit organic range? Let me start real quick. Bill gave an overview of Certara. I can do that in 30 seconds from a financial perspective too. We're a $400 million plus revenue annually company that's growing 8% to 10% on a reported basis. As Bill mentioned earlier, we are about 50/50 software services. Software organically is growing 6% to 8%. Our margin, our profitability profile is in the low 30s from an EBITDA margin perspective. I think, coming back to your question, Joe, what's going to catalyze additional growth, particularly when we think about the software results, it's the investments that we're making. We've talked a lot about our offerings and some of the offerings to come, whether it be QSP or the AI offerings. We are rolling out three new software products inside of 2025. What we're assuming is that even if the end markets remain unchanged, we expect the company to grow additionally, accelerate growth off of our organic results this year, moving into next year by making those investments. Of course, over the longer-term time horizon, as we see end markets normalize, we fully expect the company to be able to grow on a long-term basis in the 15% that it once did in history. Are renewals the big event to bring in a lot of that new decision, the new spending? I think you have some big enterprise renewals coming up. Maybe how those are looking? Yeah, those are looking good. We still have confidence in the plan in the back half of the year, which to your point, Joe, was reliant on some tier I software renewals, and we're seeing those play out as we expected. It's not just the renewals, but as we have these additional product offerings, we're also going to be able to find expansion. Part of our overall business model is to land and expand. We have many, many customers. In fact, in the tier I category, most all of the world's large biopharma companies are already our customers. It's a matter of adoption and expansion. We do that through the new product offerings that we have. OK, that's great. We're out of time. Please join me in thanking Bill and John.