MindWalk Holdings Corp. (HYFT) Earnings Call Transcript & Summary

All right. Just some preliminary comments here, everyone. Thank you, firstly, for joining us. Jennifer will be presenting and discussing the business. I'll be asking some questions, and we're digging into it a bit. And then at the end, we are taking questions from the audience. We have received many questions already. We will go through those first. And then we will go through the questions that the audience puts in there probably at the bottom of your screen there is a Q&A button type in your questions. We've addressed it, we'll skip over it, and we'll try to get through to everyone. And then with that, I'm now going to start the official presentation, Jennifer, are we good? all right. And here we go. Good day. We've got CEO, Jennifer Bath, of ImmunoPrecise Antibodies joining us today. ImmunoPrecise is an antibody development company using biology, modeling and artificial intelligence. It's Wednesday, July 19, and I'm Martin Gagel with Market Radius Research. Please remember this is neither a recommendation nor investment advice, we're here to learn about the company. Jennifer, thank you very much for joining us. We've got a big and excited audience online with us. I am new to the ImmunoPrecise story. So for myself and others new to the story, Jennifer is going to give an overview of the company and the opportunities, and then we'll dig into the details and take questions from the audience. So Jennifer, again, thank you very much for joining us. Please introduce yourself and tell us a bit about ImmunoPrecise.

Wonderful. Well, thank you, Martin, and hello, everyone. My name is Jennifer Bath and I am the CEO of ImmunoPrecise Antibodies, also known as IPA for short. We are a publicly traded company on the NASDAQ Global Market under the ticker IPA. And at IPA, we bring about 30 years of innovative antibody technologies and platform development, together with patented AI-driven technologies used to identify meaningful patterns that provide unique insights into biology, disease and medicine. We are reshaping computational systems biology and setting the stage for a future where AI is a predictive force in antibody discovery and development, driving novel insights using multimodal analysis within a single framework. In other words, collecting insights and identifying hypotheses by analyzing all different types of data in one single place. I'll share on this webinar areas in which LENSai technologies from our subsidiary, BioStrand are shaping this transformation in drug discovery and in analytics and explore the potential of our patented technology. So before we dive into the details, as always, I would just like to establish that any forward-looking statements made during the presentation are based on our best knowledge and our current understanding at this moment in time, it's really important to keep in mind, of course, that they should not be used as the sole basis for investment decisions without conducting your own research. All right. So advancements in science have really ushered off into this era where we are just steadily decoding really what is a complex language of biology. This capability grants us the ability to understand life's intricate details at actually a molecular level. And this is -- it's not a futuristic vision, it's actually something that we're able to do here and now, and much of this is made tangible by the work at IPA subsidiary, BioStrand. So this innovative blend of artificial intelligence and then biology is really beginning to revolutionize the field of computational systems biology. It's also paving new route for a therapeutic antibody discovery and development as well as data organization and data analysis. IPA really stands at the forefront of this progress driving major advances in technology. Throughout this webinar, we're going to delve into some of the inner workings of our unique capabilities, and we're going to provide you with substantiated evidence of the novelty of our capabilities and offerings and then there are potentially transformative impact on the development of novel therapy. So you're probably all aware of the fact that the world has just in abundant and diversity of just biological data, including genomic data, proteomic data and any other type of omic data that is out there. And this wealth of information has the potential to unlock a very deep understanding of life and of diseases and disease processes. However, here's the catch, this data oftentimes exists in isolation, and it's scattered across various different data sets. For instance, like in the scientific landscape, we're experiencing exponential growth in biological data. It's like just a never-ending stream of information that's coming at us. And each data set is created by different researchers and different institutions that ends up getting stored in different formats. It gets locked away in separate silos. And it's like trying to solve a puzzle when all of the pieces of the puzzle are scattered across different rooms. So the fragmentation is actually what ends up causing a huge challenge for researchers that want to collectively use and analyze this mass data. So to illustrate even within a single modality such as just text, the data can be stored in different silos. For instance, you might come across an abstract of a research paper PubMed but the access to the full article, you need to actually go and dig through a different source. It's like chasing after puzzle pieces that have been hidden in various places or better yet like trying to actually write a sentence in 5 different languages, it just does not work to bring this all together. And it doesn't end there, even if we look at like a single dimension. So if we just look at, for instance, like structural data, there are different formats and different ways to describe those structural formats. So even analyzing like data from a single modality becomes a really challenging task. It's like trying to make sense of the jigsaw puzzle when everyone uses different colors and shapes for their pieces. So this fragmentation is what we call the Information Integration Dilemma or IID for short. It's the challenge of bringing all of these scattered pieces of data together to form the complete picture. Without being able to integrate those data sets, we miss out on really valuable insights that could be gained by combining them. So again, back to my analogy, it's like looking at individual puzzle pieces without actually ever putting them together and seeing the whole picture that they create. So in today's discussion, what I'd like to do is delve into how we're tackling this challenge and how we're using our solution to actually solve problems. We'll explore the ways that BioStrand has actually overcome the IID with the Information Integration Dilemma. And then we'll show you how we've used this to unlock the true potential of data. So I want to start off with just some of the basic terminology to make sure that it's easy and clear to follow. So the terms dimension and modalities are key to understanding our work at BioStrand. So to illustrate, consider like Meta AI's recent announcement that they've developed a framework that can integrate text and image. Here, the text and the images represent two different modalities within a single framework. And that's fascinating yet the complex part of our field of biology is that we handle at least 3 different types of data: unstructured information, so things like the research papers would be an example; sequential data, so just reading through genetic code or other sequences; and structural data like the 3D molecules. The challenge that we face is about figuring out how to merge all of these different types of information from the text, from papers, genetic sequences, 3D molecular models into a single system to effectively extract meaningful information from that. So as I alluded to, BioStrand has solved this problem, which I just referred to as the IID. Lacking this ability historically, people have always chosen instead to focus on a single dimension or a single modality such as a natural language processing to harness all of the data in scientific information or focusing on, for instance, structural information such as AlphaFold does. However, until now there hasn't been an integrated framework where all of these modalities could be combined for analysis. So BioStrand has solved this issue. We've eliminated the need to choose between different modalities. We bridge these different dimensions and modalities with the aim of integrating, consolidating and then analyzing them using state-of-the-art AI algorithms and technology to gain those novel insights. So one interesting fact that I always come back to is when we tell experts that we're combining tech and fintech and structure, they recognize the enormity of this challenge. We're addressing the issue with our integrated intelligence platform which is based on a technology called a HYFT technology. And this unique approach brings together all of these different dimensions, and it effectively positions us at the forefront of health care innovation particularly in the exploration and discovery of scalable integration across these different modalities. So what I'd like to do now is just introduce you to these proprietary HYFT and what they're exactly about and how they help us accomplish this. So our unique and patented HYFT, also referred to as universal fingerprint. They provide the key that allows us to integrate and merge these various types of information and dimensions. We utilize the HYFTs to create a single, enormous knowledge graph. And with this graph in place, we can apply advanced AI, and we can apply deep learning technologies to assist experts in their decision-making process. We include these deep learning models based on statistics within our framework, and then we marry them with rule-based or straightforward inquiries for data analysis. As a result, our approach combines different strategy, applying a hybrid approach. The volume of data that we're dealing with is enormous and it's growing, and it makes scalability absolutely crucial in our field. Our HYFT-based model is extremely scalable, and it understands context at different levels. It can make predictions and they can distinguish between known information and what is actually new information. The core capability makes -- the core capability of HYFT is data integration. It integrates all of these different pillars within the biosphere. So sequence, structure, function, which is incredibly key and even text. So as such, we're able to integrate new information or models as they emerge in real time. This core capability allows us to actually stay ahead of competitors.

So just -- so your -- initially, what you do is you collect data from various disparate sources and you put into a big data graph, a big database and then you've got AI tools to sort of extract or understand it. But that graph, can you sort of attach other AI tools or models to it? So essentially, your initial value is in collecting and putting it into a usable form and then it's available to yourself or to clients to when it's in sort of one giant database. So now they -- once it's in that graph, they've got lots of different tools to extract it from it. It's the assembly of the data that is key to this?

Yes. So actually, there's a number of different ways to use it, and I love this question because we are creating proprietary databases using data that we're generating at IPA and in addition to that, the information that we have contained within our knowledge graph, which again is enormous. It has over 25 billion association within that knowledge graph is all based on information that has been collected and screened and also related back specifically to these HYFTs. And that information comes from all different sources. And again, it can be any type of data. But a lot of that is based on external databases that we're able to access and then also the internal information that we put in. And so when it comes to analyzing information for clients, we can do a number of different things. We can actually use client proprietary data which we don't store, we don't keep. They don't need to worry about us using that data later in order to analyze that data using our platform and extract insights that they wouldn't otherwise be able to see because as we push out into the LENSai and it analyzes all of this information from various databases and again, our own internal information, it's able to do so across any, again, any modality in that one single framework. So you're extracting these insights that, yes, that a client would never able to be able to collect in and to analyze and then you're getting these insights pertaining to what it is that they're researching. And yes, that knowledge graph connects all of this information. And as I'll talk about here today that HYFTs now enabled this to happen at almost an instantaneous speed and with very little energy because all of the information that actually gets input, not only is searching really all of these different modalities, but it's searching it through our patented HYFTs, which go out and find and extract the meaningful information only that come back to create this knowledge graph, which allows you to look at the connections, to identify how things are related. And also, really importantly, which I don't think we'll have time to get into today happens in a white box manner, which means you have complete traceability back to any of this information. If you're drawing a conclusion unlike, for instance, like Chat GPT and other generative AI algorithms, where it happens in a black box, you have no idea how they came to a conclusion. you have no idea if it's accurate. You can't track it back, you can't pull that as a research to validate its accuracy. We have done the exact opposite here. And so it's kind of multiple different components of what makes this really unique for that type of analysis.

All right.

Yes, it's really fun.

It is.

All right. So no -- so back to the HYFT-based technology, it's been compared, and I really like this comparison to a database [schema], really, that's generic enough to adapt to improvement in underlying technologies and then assure that we stay ahead. So as you mentioned, does it plug into other platforms? Can it be used with other platforms? Absolutely, it can and in part all of our unique capabilities to other platforms that they're connected. And our technology in and of itself is so scalable that it can handle the vast amounts of information out there in the biosphere. And then as I alluded to, it does it incredibly rapidly because of the HYFTs that are at the core of the software.

Are you using your own server farm? Or are you -- is this running on Amazon or Azure? Or how does the actual physical infrastructure work?

Yes, so we do store information on -- we do store data on AWS. However, our proprietary HYFTs and our proprietary code is actually not actually connected to any accessible database, and it's not actually live connected to the internet is something that we briefly put in to extract what we need to need immediately and then unplug so from a safety perspective.

Okay.

Yes. So I think a way to maybe also for -- because I realize some of this actually gets a little bit deeper into the complexity of data analytics. So a way to make it a little bit clearer sometimes is if we just take a look at the life of like an everyday researcher. So everyday researcher in the laboratory, whether it's an academic institution or a commercial company anywhere from small biotech to pharma. If they want to explore any certain type of disease, target, for instance, they need structural information, then they would need to go and access a very specific database for that. And if they need sequence information, then they'd have to switch to another tool. And if they want the latest inside, they would actually need to go yet again to another tool. This is actually how we did it 20 years ago when I was in graduate school, and it hasn't changed much. So that sounds a lot. And it's -- these are all separate tools and even for the most basic of research. And so this back and forth that ends up happening between databases and data types and different models is incredibly tight consuming. It slows down research at every stage. And that's what's unique about our LENSai software. Our platform is unique in that it can oversee all of the connections between these modalities right there in that single framework. So another way to look at our HYFTs is like they are -- they're like words and text. They provide meaning, they provide context and they provide ranking information. So also, if you think of basic generative AI programs, they're not able to do that. They can't tell you if a particular word is referring to a gene or a protein or what exactly its relationship is to the context around it. So like Google's PageRank, we can identify which patterns are more informative than other patterns. And this capability allows us to create massive amounts of data. We can eliminate background noise, and we can identify only what is most important for the research at hand. And we can capture all these relations from all of these different modalities using information that we can then go back and use to support wet lab discovery, which is an extremely important part of what we do at IPA as well. So we are now combining decades of scientific expertise in antibody discovery with the massive potential of AI. So our laboratory technologies, they are proven. They have successfully produced antibodies that have gone on to the clinic, supporting various pharma companies in reaching their goals. Our laboratory technologies have now been used for 19 out of the top 20 pharmaceutical companies and literally hundreds and hundreds of clients. And now we're taking that wealth of experience, and we're integrating it with our LENSai technology. And it's not just about gathering high-quality data, it's not -- this is just not the singular focus, we're making that data work smarter. So that's where our patented HYFTs really come into play with the wet lab. The HYFTs are unique integrators that facilitate a 2-way conversation between our lab and then the AI. And our lab provides the insights to the HYFTs and then return the HYFTs guide our laboratory research. And we refer to this as an active learning loop or sometimes you'll hear us refer it to it as HYFT enrichment, the HYFTs are constantly being enriched to better connect with each little pattern means with regard to structure and functional relationships. So if you imagine the exchange of insights between our LENSai software and biological systems, each new discovery adds depth to the previous ones, paves way for more precise and more efficient drug discovery and it's a synergy between knowledge and technology that just constantly is pushing the boundaries of what we're able to do. So our goal is driving the growth of our business by offering a compelling and also a very unique value proposition. Our cutting-edge and in silico technologies, combined with our award-winning wet lab expertise including our well-known function first B-cell platform, creates a powerful combination that revolutionizes how we interpret data, and that's really a key component. So there are a lot of different ways right now that we're using LENSai actively to solve various challenges for clients. We have two publicly released examples of recent real-world applications of LENSai and this platform, and how we begin to leverage our IID solution and the ultra-scalability within our software to solve complex challenges for clients. So these data provide early insights into the versatility of LENSai and they give you a glimpse into the impact that they can have on BioStrand's business, and even their early applications have already enabled the launch of unparalleled capabilities in our contract research and they've played a role in -- a very crucial role in advancing an active clinical trial. So I'll share these two examples of case studies with you. All right. So the first case study that I'm going to share with you is a little bit more in depth and a little bit more technical, but I'm going to break that down for you as you go and we've put a lot of detail in here because it's got these -- a lot of different components that are really at the core of our LENSai software, but also it's one that we're really feeling a reverberation in our market as we've put this out there over the last 3 months and clients begin to understand how it can really impact their programs in terms of speed and cost and in quality. So this case study here underscores really our initiative to transform what is known as the transgenic animal market with an integrated digital platform for what is known as humanization. And humanization is the process of taking an antibody and making it more human-like, so that the therapy doesn't have any unintended side effects when it's administered to a patient. So this service features our LENSai-driven immunogenicity workflow that has a very unique advantage over our competitors. It uses the scalability of LENSai to analyze the entire proteomes of multiple species, providing insights into what evolution declares is like human-in-nature versus nonhuman-in-nature. And it provides more thorough coverage of antibody sequences under analysis compared to competitors, which is really important because any slight change in a sequence can lead to an unintended negative clinical effect. So recent or rodent models, they have really always played a role in antibody discovery, and for these antibodies to be used safely in humans, they must either originate from these transgenic animals, because these transgenic animals have been genetically engineered to produce human antibodies. So they're human in nature already, so they're going to be safer or the antibodies need to come directly from a human or they originate from a natural animal, which has held the majority of therapies that have gone on successfully into commercialization have been discovered, where the antibody after it's discovered, it then goes through this process of what we call humanization and that process modifies the nonhuman antibody to make it safer for use in humans to make it look human, if you will. And so these transgenic animal models emerge to address what was an inefficient humanization process which included long time lines, really high cost and then also, probably more importantly, the risk that in going through the process of making that antibody human, you are actually going to lose the activity that made that molecule effective. So -- however, there is drawbacks also to the transgenic animals space. So compared to natural animals, these genetically engineered animals, they typically demonstrate a weaker immune response, and they typically have less diversity in the antibodies that they can produce, which means that overall, you're going to have fewer therapeutic candidates to choose from. And then also not to overlook that there's a lot of substantial costs associated with them. They may include royalties, licensing fees and milestone payments, and that's all being added on top of the millions of dollars of R&D expenses just to develop the drug. So IPA's new solution determines a proteins immunogenic potential by integrating artificial neural network-based binding predictions, to see if a region of an antibody is actually likely to bind to human sequences in the immune system that would indicate that, hey, that antibody, it's foreign, or that's potentially toxic, you shouldn't move forward with it. And our scalability within LENSai enables the comprehensive and very rapid identification of those dangerous antibody sequences. So our models predict that binding event and then in addition to that, it compares that to these flagged regions of the antibody to scan the entire human proteome and start pinpointing any stretches within that antibody that might be considered foreign and that might be considered dangerous. We've already calculated the immunogenicity scores of over 2,000 different antibodies. And we use that information to guide new discoveries and then improve also our method of humanization. So in this case study, if you look at the panel in the middle bottom of the screen, I see if I can point to it, right here, you can visualize also the clustering of 260 clinical antibodies that we benchmarked using our algorithms. And we segregated them based on their origin, and this also demonstrates the accuracy of our predictions that they sell into the right cell clusters with regard to origin and also with regard to clinical data that's available. However, probably the most important aspect of this workflow is the scalability and it's the rapid analysis that enables us to screen very early on in an antibody discovery program. So if a researcher is sitting there with thousands of antibodies early on in the research, which is oftentimes the case, and they need to make a decision on which ones to move forward we can very rapidly take LENSai, and we can use that to identify the antibodies that have the highest probable toxicity. We can remove those from the study very early on and then reduce the time and reduce the cost of going forward with further characterization. So it makes it easier to find the lead candidates of interest. And when that candidate is identified, we can then apply a very rapid humanization process, which is actually shown here now in the upper left-hand corner, where we utilize antibodies that, in this particular case, we went for ones that were extremely immunogenic, very toxic. They actually were identified from a chicken, which is incredibly different than a human antibody. And then we applied a very rapid humanization by making minimal changes to produce a safer version. And in this particular case, we started with one antibody pair, and so it's a heavy chain you see here at the top of this graph. These are -- antibodies are made up of 2 chains, a heavy chain and a light chain. We started with one parental heavy chain and one parental light chain and then what we did was we made minimal mutations to these, and then we rapidly identified the best therapeutic clones here. Four different versions, new versions of each of those that were the least immunogenic and the most human in nature. And in this particular study, we identified 2 different clones that you'll see boxed here in orange, H2 and then also L3 and then using data from another experiment, which you'll see here in the lower left-hand pocket here, we were able to actually demonstrate that we retained the desirable properties of the original parental clones. So in summary, we went through this process, we can go through this process of screening down all of these antibodies to find the best leads but when we get to the best leads, we can humanize them very rapidly. This analysis tells us, "Hey, these ones are the most human. These are the ones you should look to move forward, but did we retain the function? Remember, that's one of the biggest challenges in the process of humanization. And yes, we can demonstrate that we can also retain the function of these particular antibodies. So this process in and of itself is extremely rapid, it's extremely scalable, and it costs a fraction of a typical humanization program. And since first unveiling this publicly, less than 3 months ago, we've already made significant improvements in advancements. We've created new opportunities for our clients to produce safer, faster and cheaper antibodies on their way to the clinic. And at this time, we've actually queued around 10 different programs where immunogenicity and/or humanization have been added to quotes or statement of work just since unveiling this. And overall, in BioStrand offerings, we've now generated up to about $790,000 in BioStrand-specific service and quotes with over 80% of those being issued in the last 6 weeks. So our trajectory is pretty clear.

So Jennifer? So how it roughly works, there's a drug company developing something and they need some antibodies, they've identified one out of a chicken that looks like it has some attributes that could be good for us. Traditionally, they would do the transgenic mode and sort of a lot of trial and error expensive and, I guess, do animal trials and so forth? Or they can contract out to IPA and using your hybrid of your -- in silico, I guess, the term is as well as some wet lab work, you turn it around much more rapidly and presumably with a much better or a smaller, more candidate list of antibodies that have probabilistically be more likely to do the job before doing further trials on it. Did I?

Yes. That is a very good description of one of the paths that this can go. So if we take just a little bit of a broader look at what actually happens in our industry, I mean, clients can really start anywhere in a program. So they can certainly start with transgenic animals. But again, what we have found is they just don't have the same diversity or the same type of quality immune response as an animal that hasn't been genetically engineered, but they can certainly start there. And in doing so, they may not have to go through this humanization process but what we find is the quality of the lead candidates is lower. Now interestingly, what has also been revealed this year is that in several cases that have been brought to us, people are still getting lead candidates out of transgenic animals that still need to be humanized. So you go through spending all of these tens of millions of extra dollars and you're still having to go through the process it was built to avoid. One of the reasons that IPA is really successful with alternative species, whereas a lot of our competitors will offer mice and rats. And historically, that's where a lot of original antibodies were made because their antibodies kind of look very similar to humans, if you will, because their genetic relationship to humans is closer. However, you don't actually also get the same diversity of antibodies from those animals because they are so similar to humans. And so one thing that IPA has done over the last 15 years is to also branch out. We've branched out into all different types of species and our platforms are species-agnostic. So you can utilize any different species, and we are very, very specific about customizing that for you so that we can pick the right species to give you the most diverse candidates with the best probability of being successful. So we can work with the transgenics ourselves. We don't sell the transgenics, but many companies bring those to us. We can work with any different species but at the end of the day, what really matters is optimizing the diversity, optimizing the quality of the candidate and that's where we really excel. And that's where transgenic animals have struggled. So to really take the benefits of meeting a fully human sequence and then meeting the most diversity with the best candidate, we bring those together by offering this particular solution where you can use any animal that you want. You can start with the animal that's actually best for your program, and we can determine that through in silico analysis. And now you don't lose the time and the money associated with humanization and you don't lose the overall really natural diversity of an animal's immune system by going to a humanized animal. So we bring the best of both together, and so that's what enables us to turn out a better product, a faster product and a less expensive product. And so we really -- what we've really done is pinpointed the pain points for our pharma partners and then work to bring together multiple technologies to put together one comprehensive platform to solve those.

All right.

All right. So let's see I'm just going to share one more case study with you here. That's a little shorter. It's a fun case study and a little bit different in nature, in the sense that this is actually a small molecule inhibitor. So what we will share here actually really demonstrates the utilization of BioStrand's patented HYFTs and then its ability to extract valuable and actionable data from the biosphere. So what happened here was recently one of our pharma partners required the in silico identification and validation of potential inhibitors of an enzyme that they were targeting in the clinical trial. And the inhibitors needed to meet really very specific requirements for the client. And so this enzyme was already the focus of their ongoing clinical trial. And by employing a knowledge graph from our NLP and encompassing over 25 billion associations, including both structured and unstructured data, BioStrand not only successfully identified the active inhibitors but also prioritize them using our in silico docking and in silico molecular dynamic workflows. And our pharma client independently took the data that we gave them and confirmed the validity of all of the in silico inhibitors in their laboratory, including the accurate in silico-predicted rankings that we did of the relative strength of the inhibitors to bind to the target which we then actually rendered into an interactive 3-dimensional model, which you can see a static screen shot of up here on the right-hand side. And thanks to LENSai's platform, what happened here is that the screening requirements for this client reduced the number of inhibitors that needed to be screened from over 250,000 candidates to only 10. And that was really a powerful testament to the time and the energy and the money that we -- were saved by our technologies, also bringing together again in all these different modalities and one framework and different capabilities that enabled us to do something that we don't believe any other company would have been able to integrate and accomplish with their single technology. But more importantly, and I think the bigger take-home from this message is we were able to provide regulatory data for our partner in approximately one week, which enabled them to just completely move on with the rapid continuation of the clinical trial that had been underway. So it was a very successful program, and now we have a very happy client. So just to wrap things up here, having the Information Integration Dilemma, which I've described as solved, we're focusing on what our innovation enables for the future of science and health care. We believe that in the potential of our technologies to advance and pave the way for a truly personalized medicine in the stage for in silico therapeutic discovery. Our vision is the development of tailored antibodies uniquely adapted to each patient's needs. And you can see multiple ways here where we've already integrated these capabilities and these benefits into our laboratory in order to solve many of the problems and the pain points for our pharmaceutical clients. And so as I mentioned, our vision is to tailor these antibodies, make them uniquely adapted to each individual patient where physicians are then able to target the root cause of the disease by understanding the structural mutation at the molecular level, which is a core capability of LENSai and then personalizing each therapy to reduce undesirable side effects and then also to maximize the therapeutic efficacy. And just even recently, we've unveiled many different in silico technology offerings that integrate directly with our wet lab capabilities, providing considerable advantages to our clients. So we continue to unleash new potential from LENSai and to transform multiple critical sectors such as in silico, de novo, antibody discovery and development, also data organization and management, biomarker identification and early disease detection. This positions us for a significant role in the future of therapeutic antibody discovery and then by extension, as I mentioned, our goal of personalized medicines. The scalable and versatile and generic nature of our platform allows us to broaden our influence within and beyond the biotech industry without having to make any changes in our underlying technologies, offering really a vast field of commercial opportunities for IPA in the future. So we've recently laid out our specific path to commercialization and also the specific revenue streams for BioStrand and BioStrand-IPA combination revenue streams in last week's fiscal year earnings call and for anyone interested in taking a deeper dive into those revenue pathways, you can locate the transcript for this on our Investor Relations website and then just click under the sections Financials. So to wrap up this slide, we are in the process of setting the foundation for this future right now. Every action that we are undertaking today is bringing us a step closer to drastically changing how health care will look tomorrow. And the potential is vast, and we're really at the beginning of this drilling exploration, which is freely a lot of fun for IPA to watch this come to life. So in closing, choosing to invest in ImmunoPrecise Antibodies, it's not just about capital growth, it's about joining us in significant scientific and commercial adventure. You're supporting a transformative development that could reset the standards in bioinformatics and therapeutic antibody discovery and also development and a change that might remodel the medical landscape of the future. In essence, basically, your investment provides the necessary fuel to drive that innovative research and innovative business models, which could truly alter the course of health care. So that's all I've got for you today. But as you mentioned, Martin, we'll open up to Q&A, and I'm happy to answer any questions that people might have.

Sure. I just want to, with your BioStrand acquisition and -- because I just understand broadly the business models. You have -- you are -- have been historically a contract research organization and you acquired BioStrand, which is an information technology company focused on the health care or the biotech sector. So currently, you're using the BioStrand technology to enhance and your own CRO operations and services. And doing it more efficiently and effectively. Am I right on that?

That is absolutely correct. We've integrated multiple in silico capabilities into our wet lab offerings. Now what we do is we oftentimes just add them as a line item that improves the quality of the workflow, but also it improves our LENSai because it is effectively learning and enriching those HYFTs when we conduct this work. Again, without storing client data, which is unique and important to those clients. But also important for us, those in silico workflows that have been integrated into the wet lab, have very much faster revenue recognition time, so it doesn't take maybe like it would in a wet lab 3 months or 4 in the months to fully recognize the stage of a program that's oftentimes is a day or two and then extremely high profit margins over 90%, we're still working on doing the exact calculations as the research is still earlier stage. So that's absolutely true. We've integrated that into the wet lab, and then we have other ways also that we work with the in silico work. That's not a direct integration at the moment.

Will you be sort of, I saw the word SaaS on one of the last slides there. Will you be licensing it out so other wet labs either out of big pharma or other CROs can license the technology and use that as one of their tool sets so you would get that 90% gross margin, presumably on all -- on the appropriate work that they do with that as well?

Yes. So that's a great question. We get that question a lot. And yes, eventually, we do intend to offer LENSai as a SaaS model. Right now is we're just rapidly adding capabilities to LENSai, one of the core decision factors that we need to make is which modules get added to that SaaS model? And are there different modules people can purchase. Right now, that overall platform that would be rolled into a SaaS model is not ready for that, it's in an earlier stage and not in a position where we would roll it out as a SaaS model. So that's why a lot of that work is being done on the fee-for-service basis as well as we're continuing to improve it. But that is our ultimate goal that we will roll that out into a SaaS model and enable people to do all of these sorts of calculations on their own and to develop the insights into their research right there within their own institutions.

Are you able to give any sort of a time line when that would be available as a service is at like at 2023, 2024 or further any kind of without a specific date, but just rough guidance?

Yes. It's -- we've taken a look at that, especially in the last month or so. And it is a little bit difficult to pinpoint. So I'll throw that out as a little bit of a disclaimer. One of the things that impacts it greatly is the amount of capacity that we have to work on some of our other de novo in silico partnerships we have, the fee-for-service work and then actually the coding of various aspects of what will be that SaaS model. But some of our more recent analyses are looking at something maybe about 6 months from now where that platform could be unleashed. And then obviously, it would build gradually until it's being utilized by more groups and also building different modules.

Would you be able to offer it as a service to a big pharma where you go in and say, "You've got all these disparate data here and there. We're going to go in digest the data give you a giant graph and we'll charge you a service and maybe an ongoing SaaS type thing". Is that on the road map as well?

It is possible, but one of the areas where we're seeing that sort of a question emerge, and we love this because what we've realized is that even though the companies are the same that want to build new therapeutic antibodies that they're the same companies that might be sitting on 20 years of data, 30 years of data disparate data, just like we talked about the puzzle pieces in all the different rooms and they don't fit together, right? And this has turned in a real problem really for almost every pharmaceutical company we've talked about and what has really been a wonderful finding for us is that even though the people that is literally like thousands of people that we know, we've worked with and identified as key decision makers for a therapeutic antibody programs, they may not be the person that's doing all of the data organization in that company, but they know them because it's the data from their experiments and labs that they're trying to organize over here. So we've had this great opportunity to begin that dialogue with a number of different pharmaceutical companies over even mostly really in the last month. All of them have the same story, 20 years, 30 years' worth of data, it's disparate, it's off different pieces of equipment, sometimes it's in people's written notes, it's clinical trial data, structural data. Help, what do we do? And so while we were discussing, yes, that SaaS model, what we're finding right now is that people are coming to us and asking, "Can you do data organization?" and we can. We can do data organization and we organize data in a unique way, because our data organization is being done in a way that enables the rapid analysis on the backside with our HYFTs. And we can do also the data management. And so one of the things we've been working on here in the development of being really able to take on larger programs for that is really fine-tuning those models, what that looks like, what the costs are on a competitive basis. And right now, we're in active discussion with 3 different companies that would really like for us to take on and tackle at least pilot portions of data organization and management. And then obviously, a big interest for them is that once they've done that with us, they can actually tap into the power of LENSai to analyze it.

All right. Okay. Are there any other sort of models or that which I haven't asked about that -- are there permutations on monetizing this? Or is that the key potential pathway?