MindWalk Holdings Corp. ($HYFT)

Good morning, ladies and gentlemen. Thank you for joining us today for MindWalk Holdings Corp.'s Third Quarter Fiscal Year 2026 Earnings Call. MindWalk trades on the NASDAQ under the ticker HYFT. Today's call will be led by our Chief Executive Officer, Dr. Jennifer Bath; and our Chief Financial Officer, Scott Areglado. A copy of our financial statements, MD&A, is available on our website at mindwalkai.com. A replay of today's call will be available on MindWalk's Investor Relations website following the conclusion of today's call. Before we begin, please note that today's discussion includes forward-looking statements. These statements are based on current expectations and involve risks and uncertainties that may cause actual results to differ materially. For more information, please refer to our filings with the SEC and Canadian securities regulators, including our most recent Form 20-F. Unless otherwise noted, all financial figures discussed today are in Canadian dollars. I will now turn the call over to Dr. Jennifer Bath. You may begin.

Thank you very much, and good morning, everyone. This quarter, MindWalk reported its third consecutive year-over-year revenue increase and advanced 3 pipeline programs toward data readouts. In addition, we recently signed our first 1-year enterprise LensAI platform contract. I will walk you through each of those. On revenue, year-over-year, we have grown 3 quarters in a row in a market where pharmaceutical demand for AI-driven discovery is accelerating. On the commercial model, our largest enterprise AI client recently signed a 1-year LensAI platform contract, the first of its kind for us, shifting a part of our revenue from project-based to contracted and recurring. On our pipeline, Dengue, GLP-1 and influenza each have data anticipated in the near term. Please let me take those in turn. MindWalk just reported its third consecutive quarter of year-over-year revenue growth. Revenue was $4.2 million this quarter, a 52% increase from $2.7 million in the same quarter last year. MindWalk's U.S. revenue, our most important commercial market, doubled year-over-year. That growth reflects a deliberate strategic focus on the U.S. market. North America is where AI-driven discovery demand is concentrated and where the regulatory environment is actively pulling pharma toward domestic partners. We have invested in U.S. commercial presence, including business development and sales resources in the Boston and Cambridge area. Separately, we have also established biologics services operations in the Boston and Cambridge area, both reflect the same strategic direction. Our clients are pharmaceutical and biotech organizations with their own R&D capabilities. They engage us when the challenge exceeds what conventional tools can address, which brings me to the second thing I would like to highlight. Recently, our largest enterprise AI client signed a 1-year LensAI platform contract. This contract is structured as a recurring revenue model with revenues being recognized monthly. To be precise about why this matters, until now, our revenue has been primarily project-based. Clients engage us for a program we deliver, we invoice. That model produces good revenue, but it requires continuous reselling. Every quarter starts close to 0. A platform contract is structurally different. It is contracted, recurring monthly revenue that does not require reselling. It delivers value consistently, which is exactly what LensAI is designed to do. LensAI is actively being rolled out across our broader client base. The 1-year contract is one we are scaling. Now let's discuss specifically what LensAI powered by HYFT technology demonstrated this quarter. At its foundation is HYFT, our patented biological representation system that operates on the invariant functional layer of the sequence-based. Sequence-based AI tools identify patterns in surface similarity. HYFT, conversely, operates on functional architecture, the layer that governs what the molecule does, not just what it looks like. LensAI puts that capability into practice, integrated across our laboratory operations now, connecting in silico insight directly to bench-level execution. When our scientists design experiments, they identify targets and they interpret results. That capability runs through the process end-to-end. Two results this quarter illustrate what that means. First, we advanced our functional adjacency capability, the ability to identify molecules that produce the same therapeutic effect despite having very low sequence similarity. For a pharma partner, this means that LensAI can detect competitive threats and IP collision risks that conventional sequence analysis would not find. IP protection on this capability has been initiated. Second, in our influenza program, LensAI has now screened over 2,000 highly diverse influenza sequences, spanning influenza A, influenza B, avian and swine origin sequences, across all sequences analyzed, it identified a single conserved functional feature that is present in every single one, a conserved functional feature that represents a potential design target for a broadly protective immunogen. For MindWalk, Dengue is proof of concept, influenza is repeatability. Now our pipeline advancements. Dengue infects 390 million people annually. The WHO considers it a top 10 global health threat. After 60 years of research and billions of dollars of investment, the world still does not have a vaccine that reliably protects against all 4 serotypes without risk of making the disease worse. Two vaccines have reached this market, neither solved the core problem. Sanofi's Dengvaxia was restricted in 2017 after it was found to increase severe Dengue risk in seronegative patients through antibody-dependent enhancement, also known as ADE and was permanently discontinued in Brazil this year. The vaccine effectively stimulated a primary infection in seronegative recipients, priming them for enhanced disease on subsequent natural exposure. Takeda's QDENGA showed a different failure mode. It demonstrated no efficacy against serotype III in seronegative individuals and remains skewed toward Dengue-2. Takeda withdrew its FDA application in 2023. You see the problem is not generating an immune response, both of those vaccines do that. The problem is generating a balanced response across multiple serotypes and imbalanced response triggers ADE, and that makes the patient sicker. The two vaccines that have reached the market both took a tetravalent approach and hoped the immune system would respond equally, but it doesn't. Across all sequences analyzed, HYFT identified a single conserved functional constraint present in every single Dengue sequence, a potential basis for a broadly protective immunogen design. This is a discontinuous epitope. It is invisible to conventional sequence alignment tools. HYFT found it because it operates at the level of functional biological architecture, not surface sequence similarity. Instead of asking the immune system to respond equally to multiple different things, we are training it to recognize one thing, that is present in all serotypes. Balanced immunity is built into the design, not hoped for in the final outcome. Currently, rabbit immunization studies for this program are complete, binding confirmation, which is confirming that the immunized animals generated antibodies that bound to that conserved epitope is expected yet this week. Upon confirmation, we moved to multi-serotype neutralization testing with our independent collaborator. No prior program has demonstrated a single epitope immunogen generating neutralizing antibodies across all serotypes that it was immunized for. This is what neutralization data will first test. We are at this preclinical stage, but the hardest scientific questions actually get answered here. In vitro GLP-1 receptor activation was confirmed by an independent third-party assay. Results demonstrate activity relative to semaglutide, a market-leading GLP-1 therapy. We have worked with a pharma collaborator with recognized expertise in this area. They have shared what they consider important to see as this program advances. We are developing the program with that input in mind. Beyond the GLP-1 pathway itself, we have identified a dual pathway regimen, linking GLP-1 biology to a second nonoverlapping longevity pathway. We will continue to update the market as this program advances. Our influenza program is advancing on the same design logic. As of this week, we are moving toward manufacturing of the lead in silico candidate. We will update the market as that program continues to develop. U.S. revenue doubled year-over-year. a direct result of our deliberate strategic focus on North America. AI-driven biologics demand is concentrated in this market, and the regulatory environment is increasingly favorable to domestic partners. We have established biologics services operations in the Boston Cambridge area, and this strategic direction guided our decision to divest our European operations in favor of North American growth. We ended Q3 with $14.2 million in cash. The Netherlands divestiture proceeds are being deployed deliberately into commercial growth, LensAI, and its pipeline assets and our Canadian laboratory capabilities. Our team published a peer-reviewed study in biomacromolecules, the American Chemical Society Journal, in collaboration with Eindhoven University of Technology and Radboud University Medical Center. That work was grant-funded, and it demonstrates what our wet lab nanobody discovery is capable of and the great importance of this innovation. I will come back to this when I describe our B-cell Llama platform launch. This quarter, we announced results from a client-driven research engagement in which our scientists generated and validated monoclonal antibodies and interbodies capable of selectively targeting misfolded pathogenic TDP43, while leaving healthy TDP43 intact. TD43 is implicated in ALS, frontotemporal dementia and some Alzheimer's cases. Last week, we announced the launch of our B-cell Llama, a nanobody discovery platform built on single B-cell isolation from immunized Llamas. Let me explain why this matters. Bispecific and multispecific antibodies require 2 heavy chains and when those chains need two different chains, the result is an explosion of possible combinations, only one of which is the product that you actually want. That chain pairing problem has been one of the central engineering bottlenecks limiting bispecific drug development and significant capital has been invested in platforms designed to work around it. VHH nanobodies eliminate the problem by design. They carry no light chain, there is no pairing ambiguity, and because they come from a naturally matured Llama immune repertoire, they capture sequence diversity that engineered platforms structurally cannot replicate. Our peer-reviewed biomacromolecules publication demonstrates what that produces. The molecule with the strongest binding affinity in our assays delivers 0 functional activity, a construct built from the same nanobody building blocks achieved 10 to 25x greater potency in multivalent format. Function-based selection, not affinity is what matters. That is what B-cell Llama is designed to deliver. MindWalk holds commercial right to the jointly developed intellectual property from that work. B-cell Llama operates alongside our B-cell select, our existing platform with over 15 molecules advanced to the clinic. The full detail is in last week's announcement. Across our proprietary asset portfolio, GLP-1, Dengue and influenza and at the request of investors, we are working with legal and financial advisers to design structured asset-level financing vehicles that will allow investors to participate at the program level while preserving parent company equity. That work is active and progressing. I will now turn the call over to Scott.

Thank you, Jennifer, and good morning, everyone. As a note, all figures are in Canadian dollars and relate to continuing operations, unless stated otherwise. Revenue for Q3 was $4.2 million, or a 52% increase from $2.7 million in Q3 of last year. As Jennifer noted, this is our third consecutive quarter of year-over-year revenue growth. U.S. revenue doubled year-over-year, $2.6 million versus $1.3 million. The U.S. is named a strategic priority. AI-driven discovery demand is concentrated here and our commercial investments are reflected in the numbers. For the 9-month period ending January 31, 2026, our revenue was $11.4 million, as compared to $7.9 million, or a 45% increase as compared to the prior year period. Gross margin for the 3 months ended January 31, 2026, was 59% as compared to 65% in the prior year period. For the 9-month period ended January 2026, gross margin was 58%, as compared to 53%, a 5 percentage point improvement over the same period last year. Gross margin can vary depending on our mix of business. However, as we develop an increased adoption of the tools within our LensAI platform, we would expect margins to expand. Moving on to operating expenses. For the third quarter of 2026, R&D expense was $1.2 million as compared to $0.9 million for the prior year period, due to the investments in the Dengue, GLP-1 and B-cell Llama programs and ongoing LENS AI platform development. For the 9-month period ended January 31, 2026, R&D expense was $3.5 million versus $3.4 million in the prior year. Sales and marketing for the 3-month period ended January 31, 2026, was $1.8 million as compared to $1.1 million in the same period last year, reflecting our continued commercial expansion primarily in the U.S. with programs such as our expansion in the Boston area, starting to yield revenue. For the 9-month period ended January 2026, sales and marketing expense was $4.3 million, compared to $2.7 million for the 9 months ended January 2025. G&A was $3.1 million for the third quarter of 2026 as compared to $2.8 million for the third quarter of 2025. G&A expense was $9.5 million for the 9 months ended January 2026, as compared to $9.1 million for the prior year period. We expect G&A to remain flat to modest growth as we believe we have the infrastructure to support future growth. Net loss from continuing operations for Q3 2026 was $3.9 million versus $22 million in Q3 '25. Net loss in the prior year period included an impairment charge of $21.2 million. For the 9-month period ended January 2026, net loss was $11.2 million as compared to $29.7 million for the 9-month period ended January 2025, which also reflected the $21.2 million charge. We are investing ahead of revenue in commercial infrastructure, pipeline programs and platform capabilities with the expectation that these investments will yield returns. Moving on to the balance sheet. We ended the third quarter with $14.2 million in cash. Cash used in operations was $10.1 million year-to-date, consistent with our planned investments. In summary, revenue has grown year-over-year, and we have demonstrated the ability to execute. We have developed a platform and products that bring value to our customers, and we continue to innovate with programs such as our recent announcement of our B-cell Llama capability and functional adjacency. We have cash runway for operations and the capital structure to support the ongoing development of our proprietary pipeline assets. We believe this will continue to drive shareholder value. I will now return the call to Jennifer.

Thank you, Scott. Before we open for questions, I would like to leave you with this. Most AI approaches in biologics today operate on full biological sequences. They tokenize, they train, they generate. Many are powerful, and they are operating on a representation of biology that includes a great deal of noise. Evolution is a tolerant process. Most positions in a biological sequence can change without consequence. That variation fills the public databases that these models train on, a much smaller set of subsequences is in variant. They cannot change because essential biological function depends on them. These are the fingerprints that actually carry the information for life. HYFT is our patented representation of that invariant layer. No other company has the right to use these patterns. That is the foundation of a durable competitive position, because every result we generate every insight we deliver and every asset we build, rests on a biological foundation that competitors cannot replicate and is producing results. We identified a Dengue epitope conserved across all 4 serotypes. The target that 60 years of vaccinology did not find. We detected functional adjacency that sequence-based platforms miss and initiated IP protection on that capability. We screened over 2,000 influenza sequences, and found a single conserved biological feature present in every single one. Our GLP-1 candidate demonstrated activity relative to semaglutide, the market-leading GLP-1 therapy, in independent third-party in vitro testing. We launched B-cell Llama, a nanobody discovery platform anchored by peer-reviewed evidence that function-based candidate selection outperforms affinity-based selection at the molecular level. On commercial, we are scaling the enterprise platform model, additional contracted recurring platform agreements with major pharma and biotech partners building a revenue base that grows independently of any single project. On pipeline, Dengue neutralization data is our nearest-term pipeline. Dengue is proof of concept for what HYFT can do. Influenza is repeatability. Together, they make the platform case to pharma partners better than anything else that we could say. On asset financing, legal and financial advisers are engaged and structures are being designed across the proprietary portfolio. Before we open for questions, I want to leave you with this. The science is patented. The results are peer-reviewed the first enterprise contract is signed. The pipeline has meaningful data approaching. These 3 consecutive quarters of year-over-year revenue growth, U.S. revenue doubled and made a platform that no competitors can replicate. This is the MindWalk investment case. Thank you. We will now open the line for questions.

[Operator Instructions] Your first question comes from the line of Swayampakula Ramakanth with H.C. Wainwright.

This is RK from H.C. Wainwright. This is a great quarter. A lot of good stuff and really exciting days for you guys. Jennifer and Scott, in terms of the agreement that you just signed, the enterprise client agreement that you just signed on the recurring contract. I'm trying to understand what drove this group to do this. What was the primary driver? And then second part of that same question is, how many of your other project-based clients are willing to convert into this monthly recurring model, let's say, over the next 6 to 12 months?

Thank you, RK. Thanks for joining and as usual, for your thoughtful questions. So your first question, what really drove this first pharma client to go ahead and sign this contract? That's a very good question. I think I like this in particular because giving me the opportunity to explain this also gives me the opportunity to demonstrate the validation that needed to occur before a client took this type of a commitment long term with us. I do believe this is a client I've referred to anecdotally historically, 1 or 2 times, and it is a client who initially came to us, having tried multiple other companies that said that they could utilize artificial intelligence to help solve some of their problems, some of their scientific challenges. And the group was relatively dismayed. They said that it, in reality, none of those CRO partners or companies were able to turn back results that were as good as what they could do in the wet lab. And so they were apprehensive, and they were doubtful. And so when we first brought this group in, it was actually for fee-for-service work. And what we said to them is, we know you have programs that have been extremely difficult and you've worked on for over a decade. Let us take a crack at it. Let us apply LensAI to it, and if we are not successful, then you don't pay us. But we really want to show you what we can do. And we worked on that program for them, and we were successful, and they saw the outputs coming directly from LensAI. And even some applications that the MindWalk in Belgium, also known as BioStrand, built specifically for producing these outcomes in the program. They were tremendously happy with the results, and they've now contracted us, I'm not sure, somewhere between 7 to 10x in total for different programs. And LensAI has continued to successfully solve very challenging problems for them. And that is really where we earned their respect, and I think their trust for the LensAI program, and that is what really brought them to the table to negotiate the platform license as a SaaS model. And so our intent, obviously, is to leverage that experience with them to be able to bring on additional clients for those clients to understand and us also to be able to share the positive experience this group has had. Now that being said, for your second question, we're not providing specific pipeline numbers or time lines for additional contracts, but one thing that I think is really important to highlight, and maybe wasn't highlighted enough in the earnings call, is that LensAI is now actively being rolled out across our broader client base. And so all the programs we're working on, not just the programs we're working on in Belgium, where this LensAI lives, but also in Canada with all of our wet lab clients, right? So somewhere close to 750 active clients, dozens of programs running at any given time. Those results are all now finally coming back in the LensAI portal. So these groups are receiving secure login, and when they log in, they have access to this portal and they can see the applications that are in there that truly change the way they have done drug discovery historically. Now they can utilize these applications and instead of going 2, 3, 4, 5, 6 other vendors to collect information or kind of chugging through the process over the course of 18 months to 2 years they can literally take a subscription to utilize these applications beyond the base level to harness the power and get the results that they're looking for. And so being at that point in this venture is very important to our company. It's something we've built toward and worked toward. It took longer than we hoped it would to get this software into the hands of these clients, and it's now happening not just across our therapeutic clients, but clients who have contracted us really for any sort of custom antibody work. And so with regard to that, when we think about additional contracts and bringing these new clients, and that's where we're really focused, we feel we have an extremely unique situation where these clients are already onboarded. We are, in many cases, are primary vendor, but in all cases, we are a vendor that's in their system. And we've already built their trust and their respect. And so we have a very unique segue into this market with those clients.

Thanks for that detailed answer. And if I may, a second question is, this is on the asset-level financing. As -- I do understand lawyers and the investors can take a long time to come to a conclusion about anything, but how much of that are you waiting for in terms of these 4 different projects/platforms that you have thinking about the Dengue, the GLP-1, Llama, all of these, are you on the influence as well? That's the fourth one. How do you need to get to a conclusion with these groups before you move these forward? Or these are all independent of each other, and they are all moving forward?

So that is a great question. So the short answer is they're independent of one another as they move forward. A couple of things to keep in mind. When we look at financing these particular programs I think that one of the things that's easy to overlook is the fact that our program costs are not what you would expect from a traditional drug development company at this stage? And so much of our work is in silico, but also much of our in silico work, our in vitro work and even our preclinical work is either AI-driven or it's conducted in-house. So that keeps our costs meaningfully lower than a conventional pipeline of this breadth would require. And it's also one of the structural advantages that we have building on the HYFT platform. And so as a result of that, and directly in reference to your question, RK, the capital that we currently have is capital that is enough to drive us significantly forward. in these engagements. As a matter of fact, as one of the things that was detailed by Scott was the R&D expenses, which are not up significantly over last year and yet cover not only our traditional R&D and the build-out of the B-cell platform, but also covers everything we've done to date here. And so that gives you, I think, kind of a specific example of that. Now when it comes to the asset-level financing, that is something that definitely as these programs become more advanced, one of the things that we haven't talked too much about, but that we've ensured we have in place as we move forward as a professional team that has the experience in the clinical realm and the subject matter experience, which, with each of these families of viruses or the particular therapeutic or disease that we're targeting. In order to help drive this process along through the preclinical portion in the IND enabling the IND filing and the clinical readiness. And when we get to those stages, of course, then the cost does begin to increase. And so as to whether or not these portions at that stage can move forward, prior to the asset, the asset-level financing, to some extent, yes, most definitely, once again, because we do have a team set forward here with the with the internal expertise. But in addition to that, the asset-level financing is meant to support once we get to that stage. We have enough runway here that the lead time that it takes to actually get these ring-fenced should be one that enables us to bring in additional capital to support those by that time.

There are no further questions at this time. I will now turn the call back over to Dr. Bath for closing remarks.

Great. Thank you so much. All right. So really, the biggest thing that I want to say is thank you. Well, thank you all for joining us. Thank you all for supporting MindWalk. We look forward to sharing pipeline results as they become available. And we will speak with all of you on our Q4 and fiscal year-end '26 earnings call. Thank you.

This concludes today's MindWalk Holdings Q3 Fiscal Year 2026 Earnings Call. Thank you for your participation. You may now disconnect.