Mindwalk Holdings Corp. (HYFT) 2026 Q3 法說會逐字稿

Please be advised that today's conference is being recorded. 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.

敬請注意，今天的會議將進行錄音。各位女士、先生，早安。感謝各位今天參加 MindWalk Holdings Corp. 2026 會計年度第三季財報電話會議。MindWalk 於那斯達克（NASDAQ）以代號 HYFT 交易。今天的會議將由我們的執行長 Jennifer Bath 博士，以及財務長 Scott Areglado 主持。

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.

我們的財務報表與管理階層討論與分析（MD&A）可於 mindwalkai.com 網站取得。今天會議結束後，MindWalk 投資人關係網站將提供本次電話會議的重播。

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.

在開始之前，請注意今天的討論包含前瞻性陳述。這些陳述係基於目前的預期，並涉及可能導致實際結果出現重大差異的風險與不確定性。更多資訊請參閱我們向美國證券交易委員會（SEC）及加拿大證券監管機構提交的文件，包括我們最新的 Form 20-F。除非另有說明，今天討論的所有財務數字均以加幣計。

I will now turn the call over to Dr. Jennifer Bath. You may begin.

接下來我將把電話會議交給 Jennifer Bath 博士。您可以開始了。

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.

非常感謝，各位早安。本季，MindWalk 報告營收連續第三個季度年增，並推進 3 項管線計畫朝近期數據讀出邁進。此外，我們近期簽下首份為期 1 年的企業級 LensAI 平台合約。我將逐一為各位說明。

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.

在營收方面，在藥廠對 AI 驅動研發探索需求加速的市場環境下，我們已連續 3 個季度實現年增成長。在商業模式方面，我們最大的企業級 AI 客戶近期簽署了為期 1 年的 LensAI 平台合約，這是我們首度簽下此類合約，使我們部分營收由專案制轉為簽約制與可重複、經常性收入。

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 US revenue, our most important commercial market, doubled year-over-year. That growth reflects a deliberate strategic focus on the US market.

在管線方面，登革熱、GLP-1 與流感三項計畫的數據預期將在近期公布。請讓我依序說明。MindWalk 剛剛報告營收連續第三個季度年增成長。本季營收為 420 萬美元，較去年同期的 270 萬美元成長 52%。MindWalk 的美國營收——我們最重要的商業市場——年增翻倍。這項成長反映了我們對美國市場的刻意且策略性的聚焦。

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 US commercial presence, including business development and sales resources in the Boston and Cambridge area.

北美是 AI 驅動研發探索需求最集中的地區，同時監管環境也正積極推動藥廠與本土合作夥伴合作。我們已投資於美國的商業布局，包括在波士頓與劍橋地區配置商務拓展與銷售資源。

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.

另外，我們也在波士頓與劍橋地區建立了生物製劑服務營運據點，兩者皆反映相同的策略方向。我們的客戶為具備自身研發能力的製藥與生技組織。當挑戰超出傳統工具所能處理的範圍時，他們便會與我們合作，這也帶出我想強調的第二點。

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.

近期，我們最大的企業級 AI 客戶簽署了為期 1 年的 LensAI 平台合約。此合約採經常性收入模式，收入將按月認列。更精確地說明其重要性：截至目前，我們的營收主要來自專案制。客戶委託我們交付某個計畫，我們開立發票。此模式能帶來不錯的營收，但需要持續不斷地重新銷售。

Every quarter starts close to 0.

每一季幾乎都是從接近 0 開始。

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.

平台合約在結構上則不同。它是已簽約的、每月經常性收入，不需要重新銷售。它能持續交付價值，這正是 LensAI 的設計初衷。LensAI 正積極在我們更廣泛的客戶群中推廣部署。這份 1 年期合約是我們正在擴大規模的模式。

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.

接下來我們具體談談本季由 HYFT 技術驅動的 LensAI 展示了什麼。其核心是 HYFT——我們已取得專利的生物表徵系統，運作於以序列為基礎之不變功能層。以序列為基礎的 AI 工具會在表面相似性中辨識模式。相對地，HYFT 運作於功能架構層，也就是決定分子「做什麼」而不僅是「看起來像什麼」的那一層。LensAI 將此能力落地實作，現已整合至我們的實驗室營運流程中，把 in silico 洞見直接連結到實驗台（bench）層級的執行。

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.

當我們的科學家設計實驗時，他們會辨識標的並解讀結果。這項能力貫穿整個端到端流程。本季有兩項成果可說明其意義。第一，我們推進了「功能鄰接」能力，也就是能辨識出即使序列相似度很低、但仍可產生相同治療效果的分子。

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.

對藥廠合作夥伴而言，這代表 LensAI 能偵測到傳統序列分析找不到的競爭威脅與智慧財產（IP）衝突風險。我們已啟動針對此能力的 IP 保護程序。

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.

第二，在我們的流感計畫中，LensAI 目前已篩選超過 2,000 條高度多樣化的流感序列，涵蓋甲型流感、乙型流感，以及禽源與豬源序列；在所有分析的序列中，它辨識出一個在每一條序列中都存在的單一保守功能特徵——此保守功能特徵代表一個可用於設計廣泛保護性免疫原的潛在設計標的。對 MindWalk 而言，登革熱是概念驗證，流感則是可重複性。

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.

接下來談我們的管線進展。登革熱每年感染 3.9 億人。世界衛生組織（WHO）將其列為全球十大健康威脅之一。在 60 年研究與數十億美元投入之後，世界仍缺乏一種能可靠保護所有 4 種血清型、且不會增加病情惡化風險的疫苗。已有兩款疫苗進入市場，但都未解決核心問題。

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.

賽諾菲（Sanofi）的 Dengvaxia 於 2017 年被限制使用，因為發現其會透過抗體依賴性增強作用（antibody-dependent enhancement，亦稱 ADE）提高血清陰性患者發生重症登革熱的風險，並於今年在巴西永久停用。該疫苗實際上在血清陰性接種者體內誘發了類似初次感染的反應，使其在後續自然暴露時更容易出現加重的疾病。

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.

武田（Takeda）的 QDENGA 則呈現不同的失敗模式。它在血清陰性個體中對第三血清型（serotype III）未顯示有效性，且保護效果仍偏向登革熱第 2 型。武田於 2023 年撤回其向 FDA 提交的申請。問題不在於是否能引發免疫反應，這兩款疫苗都能做到。問題在於要在多個血清型之間產生均衡反應；不均衡的反應會觸發 ADE，進而使患者病情更嚴重。

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.

目前已上市的兩款疫苗都採用四價（tetravalent）策略，並寄望免疫系統能平均反應，但事實並非如此。在所有分析的序列中，HYFT 辨識出一個在每一條登革熱序列中都存在的單一保守功能約束，可能成為設計廣泛保護性免疫原的基礎。這是一個不連續表位（discontinuous epitope）。傳統序列比對工具無法看見它。HYFT 能找到它，是因為它運作於功能性生物架構層級，而非表面序列相似性。

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.

我們不是要求免疫系統對多種不同事物做出同等反應，而是在訓練它去辨識一個在所有血清型中都存在的單一目標。免疫平衡已內建於設計之中，而不是寄望於最終結果。目前，本計畫的兔子免疫研究已完成；預計最快在本週即可取得結合確認結果，以確認受免疫動物產生了可與該保守表位結合的抗體。

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.

一旦確認完成，我們便與獨立合作夥伴推進至多血清型的中和測試。過去沒有任何計畫證明：單一表位免疫原能在其所免疫的所有血清型上產生具中和作用的抗體。這正是中和數據首先要驗證的重點。我們目前仍處於臨床前階段，但最困難的科學問題其實就是在這裡得到解答。

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.

體外 GLP-1 受體活化已由獨立第三方檢測確認。結果顯示其活性可與司美格魯肽（semaglutide）相比，後者為市場領先的 GLP-1 治療方案。我們已與在此領域具公認專長的藥廠合作夥伴合作。他們也分享了在本計畫推進過程中，他們認為需要看到的關鍵要點。我們正依據這些意見來開發本計畫。

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.

除 GLP-1 路徑本身之外，我們已辨識出一套雙路徑療法方案，將 GLP-1 生物學與第二條不重疊的長壽路徑連結。我們將隨著本計畫推進持續向市場更新。我們的流感計畫也正依循相同的設計邏輯推進。截至本週，我們正朝向製造首要的 in silico 候選物邁進。隨著該計畫持續發展，我們也會向市場更新。

US 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.

美國營收年增一倍，這是我們刻意將策略重心聚焦於北美的直接成果。AI 驅動的生物製劑需求集中於此市場，且監管環境也愈來愈有利於本土合作夥伴。我們已在波士頓—劍橋地區建立生物製劑服務營運據點；此一策略方向也引導我們決定出售歐洲業務，轉而支持北美成長。

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.

我們在第三季末的現金為 1,420 萬美元。荷蘭資產出售所得正被審慎投入於商業成長、LensAI 及其管線資產，以及我們在加拿大的實驗室能力。

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.

我們團隊與埃因霍溫理工大學（Eindhoven University of Technology）及拉德堡德大學醫學中心（Radboud University Medical Center）合作，在美國化學學會期刊《Biomacromolecules》發表了同儕審查研究。該研究由補助金資助，展示了我們濕實驗室奈米抗體發現能力的可行性，以及此項創新的重大重要性。在我介紹我們 B-cell Llama 平台的推出時，會再回到這一點。

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.

本季，我們公布了一項由客戶驅動的研究合作成果：我們的科學家生成並驗證了單株抗體與 interbody，能選擇性鎖定錯誤折疊的致病性 TDP43，同時保留健康的 TDP43 不受影響。TDP43 與 ALS（肌萎縮性側索硬化症）、額顳葉失智症以及部分阿茲海默症病例相關。

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.

上週，我們宣布推出 B-cell Llama：一個建立在自免疫羊駝進行單一 B 細胞分離之上的奈米抗體發現平台。我來說明這為何重要。雙特異性與多特異性抗體需要兩條重鏈；當這些鏈必須是兩種不同的鏈時，可能的組合會爆炸性增加，而其中只有一種才是你真正想要的產品。這個鏈配對問題一直是限制雙特異性藥物開發的核心工程瓶頸之一，市場也已投入大量資本於各種平台以設法繞過此問題。

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.

VHH 奈米抗體在設計上就消除了這個問題。它們不帶輕鏈，因此不存在配對歧義；而且由於其來源為自然成熟的羊駝免疫庫，能捕捉到工程化平台在結構上無法複製的序列多樣性。我們經同儕審查的《Biomacromolecules》論文展示了這會產生什麼結果。在我們的測試中，結合親和力最強的分子卻呈現 0 的功能活性；而以相同奈米抗體構件組成的另一種構築體，在多價形式下達到高出 10 至 25 倍的效力。

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.

真正重要的是以功能為基礎的篩選，而不是親和力。這正是 B-cell Llama 的設計目標。MindWalk 擁有該項共同開發智慧財產的商業權利。B-cell Llama 與我們既有的 B-cell select 平台並行運作；該既有平台已有超過 15 個分子推進至臨床階段。完整細節已載於上週的公告中。

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.

針對我們自有資產組合（GLP-1、登革熱與流感），並應投資人要求，我們正與法律與財務顧問合作，設計結構化的資產層級融資工具，使投資人能在計畫層級參與，同時保留母公司股權。這項工作正在進行並持續推進。

I will now turn the call over to Scott.

接下來我把電話會議交給 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 Jim noted, this is our third consecutive quarter of year-over-year revenue growth. US revenue doubled year-over-year, $2.6 million versus $1.3 million.

謝謝你，Jennifer，各位早安。提醒一下，除非另有說明，所有數字均以加幣計算，且與持續營運業務相關。第三季營收為 420 萬美元，較去年第三季的 270 萬美元成長 52%。如 Jim 所提，這是我們連續第三個季度實現營收年增。美國營收年增一倍，為 260 萬美元，相較於 130 萬美元。

The US 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.

美國被列為策略優先市場。AI 驅動的發現需求集中於此，我們的商業投資也反映在數字上。截至 2026 年 1 月 31 日止的 9 個月期間，我們營收為 1,140 萬美元，相較於前一年度同期的 790 萬美元，年增 45%。

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.

截至 2026 年 1 月 31 日止三個月的毛利率為 59%，相較於前一年度同期為 65%。截至 2026 年 1 月止 9 個月期間，毛利率為 58%，相較於 53%，較去年同期提升 5 個百分點。毛利率可能會因我們的業務組合而有所波動。不過，隨著 LensAI 平台內工具的採用率提高，我們預期毛利率將擴大。

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 US with programs such as our expansion in the Boston area, starting to yield revenue.

接著談營運費用。2026 年第三季研發費用為 120 萬美元，相較於前一年度同期的 90 萬美元，主要因投資於登革熱、GLP-1 與 B-cell Llama 計畫，以及持續的 LENS AI 平台開發。截至 2026 年 1 月 31 日止 9 個月期間，研發費用為 350 萬美元，前一年度為 340 萬美元。截至 2026 年 1 月 31 日止三個月期間，銷售與行銷費用為 180 萬美元，相較於去年同期的 110 萬美元，反映我們持續擴大商業布局，主要在美國；例如我們在波士頓地區的擴張已開始帶來營收。

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.

截至 2026 年 1 月止 9 個月期間，銷售與行銷費用為 430 萬美元，相較於截至 2025 年 1 月止 9 個月期間的 270 萬美元。2026 年第三季一般及行政（G&A）費用為 310 萬美元，相較於 2025 年第三季的 280 萬美元。截至 2026 年 1 月止 9 個月期間，一般及行政（G&A）費用為 950 萬美元，相較於前一年度同期的 910 萬美元。

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.

我們預期 G&A 將維持持平至溫和成長，因為我們相信我們已具備支援未來成長所需的基礎設施。2026 會計年度第三季（Q3 2026）持續營運之淨損為 390 萬美元，相較於 2025 會計年度第三季（Q3 '25）的 2,200 萬美元。前一年度同期的淨損包含 2,120 萬美元的減損費用。

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.

截至 2026 年 1 月止的 9 個月期間，淨損為 1,120 萬美元，相較於截至 2025 年 1 月止的 9 個月期間之 2,970 萬美元；後者同樣反映了該筆 2,120 萬美元的費用。我們在商業化基礎設施、產品管線計畫與平台能力上超前於營收進行投資，並預期這些投資將帶來回報。

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.

接著談資產負債表。我們在第三季末的現金為 1,420 萬美元。年初至今營運使用現金為 1,010 萬美元，符合我們規劃中的投資。

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.

總結而言，營收較去年同期成長，而我們也展現了執行能力。我們已開發出能為客戶帶來價值的平台與產品，並持續透過各項計畫創新，例如我們近期宣布的 B-cell Llama 能力與功能鄰接（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.

我們具備足以支應營運的現金續航力，以及可支援我們自有產品管線資產持續開發的資本結構。我們相信這將持續推動股東價值。

I will now return the call to Jennifer.

我現在把電話交回給 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.

謝謝你，Scott。在我們開放提問之前，我想留給各位這段話。目前生物製劑領域多數 AI 方法都是在完整的生物序列上運作。它們進行分詞（tokenize）、訓練、生成。其中許多方法很強大，但它們所運作的生物學表徵包含大量雜訊。

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.

演化是一個具容忍度的過程。生物序列中的大多數位置都可以改變而不造成後果。這些變異充斥於這些模型所訓練的公共資料庫中，而其中只有更小的一組子序列是不變的（invariant）。它們不能改變，因為關鍵的生物功能仰賴它們。這些才是真正承載生命資訊的指紋。

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.

HYFT 是我們已取得專利、用來表徵該不變層的方式。沒有其他公司有權使用這些模式。這就是可長期維持的競爭優勢之基礎。因為我們產出的每一個結果、提供的每一項洞見、打造的每一項資產，都建立在競爭對手無法複製且正在產生成果的生物學基礎之上。

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.

我們辨識出一個在所有 4 種血清型中皆保守的登革熱表位（epitope）。這是 60 年疫苗學都未能找到的標的。我們偵測到以序列為基礎的平台所忽略的功能鄰接，並已就該能力啟動智慧財產保護。我們篩選了超過 2,000 條流感序列，並在每一條序列中都找到一個單一且保守的生物特徵。

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.

我們的 GLP-1 候選藥物在獨立第三方體外（in vitro）測試中，展現出相對於 semaglutide（市場領先的 GLP-1 治療）的活性。我們推出了 B-cell Llama——一個奈米抗體（nanobody）發現平台，其基礎是經同儕審查的證據，顯示在分子層級上，以功能為基礎的候選物篩選優於以親和力為基礎的篩選。

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.

在商業化方面，我們正在擴大企業平台模式，與大型製藥與生技合作夥伴簽訂更多具合約約束的經常性平台協議，建立一個可不依賴任何單一專案、而能獨立成長的營收基礎。

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.

在產品管線方面，登革熱中和（neutralization）數據是我們近期最接近的管線里程碑。登革熱是 HYFT 能力的概念驗證（proof of concept）。流感則是可重複性（repeatability）。兩者結合，對製藥合作夥伴所呈現的平台論述，比我們能說的任何話都更有說服力。

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.

在資產融資方面，我們已聘請法律與財務顧問，並正在針對自有資產組合設計各種架構。在我們開放提問之前，我想留給各位這段話。科學已取得專利。成果已通過同儕審查，第一份企業合約也已簽署。產品管線也即將有具意義的數據出爐。

These 3 consecutive quarters of year-over-year revenue growth, US revenue doubled and made a platform that no competitors can replicate. This is the MindWalk investment case.

連續 3 個季度的年對年營收成長、美國營收倍增，並打造出一個競爭對手無法複製的平台。這就是 MindWalk 的投資論點。

Thank you. We will now open the line for questions.

謝謝。我們現在開放提問。

(Operator Instructions) Swayampakula Ramakanth; H.C. Wainwright.

(接線員指示) Swayampakula Ramakanth；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 agreement -- the enterprise client agreement that you just signed on the recurring contract.

我是 H.C. Wainwright 的 RK。這是一個很棒的季度。有很多好消息，對你們來說真的是令人振奮的時刻。Jennifer 和 Scott，關於你們剛簽署的協議，也就是你們剛簽下的企業協議——那份具經常性合約的企業客戶協議。

I'm trying to understand what drove this group to do this? What is 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?

我想了解是什麼驅使這個團隊做出這個決定？主要驅動因素是什麼？同一個問題的第二部分是，在接下來 6 到 12 個月內，你們其他以專案為基礎的客戶，有多少願意轉換成這種按月經常性模式？

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.

謝謝你，RK。謝謝你加入，也一如往常地提出深思熟慮的問題。所以你的第一個問題，究竟是什麼促使第一家製藥客戶願意往前走並簽下這份合約？這是個非常好的問題。我特別喜歡這個問題，因為讓我有機會解釋，也讓我有機會說明：在客戶願意對我們做出這種長期承諾之前，必須完成哪些驗證。

I do believe this is a client I've referred to anecdotally historically, 1 or 2 times. And then as 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.

我相信這是我過去曾以軼事方式提過 1、2 次的客戶。這位客戶最初找上我們之前，已經嘗試過多家聲稱能利用人工智慧協助解決其問題、一些科學挑戰的公司。而該團隊相當失望。他們表示，實際上那些 CRO 夥伴或公司都無法回傳比他們在濕實驗室（wet lab）自己做得更好的結果。

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.

因此他們抱持疑慮，也不太相信。所以我們第一次把這個團隊帶進來時，實際上是先做按服務收費（fee-for-service）的工作。我們對他們說，我們知道你們有一些極其困難、而且已經投入超過十年的計畫。讓我們來試試看。

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.

讓我們把 LensAI 應用在上面；如果我們沒有成功，那你們就不用付費。但我們真的很想讓你們看到我們能做到什麼。我們為他們做了那個計畫，而且成功了；他們看到了直接由 LensAI 產出的結果。甚至還有一些由比利時的 MindWalk（也就是 BioStrand）專門打造、用於在該計畫中產生這些成果的應用。他們對結果非常滿意，現在已與我們簽約，我不確定，總計大概是 7 到 10 倍的規模、涵蓋不同計畫。而 LensAI 也持續成功地為他們解決非常具挑戰性的問題。

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.

而那正是我們贏得他們尊重、以及我認為他們對 LensAI 計畫信任的關鍵所在；也正因如此，他們才願意坐下來，就以 SaaS 模式進行平台授權展開談判。

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 also to be able to share the positive experiences group has had.

因此，我們的意圖很明確，就是運用與他們合作的這段經驗，來爭取更多客戶，讓這些客戶能夠理解，同時也能分享該團隊所獲得的正面經驗。

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.

話雖如此，針對你的第二個問題，我們不會提供新增合約的具體管線數字或時間表；但我認為有一點非常重要、也許在財報電話會議中沒有被充分強調的是：LensAI 現在正積極在我們更廣泛的客戶群中推廣部署。因此，我們正在進行的所有專案，不僅是比利時這個 LensAI 所在地的專案，也包括加拿大所有濕實驗室（wet lab）客戶的專案，對吧？也就是大約接近 750 位活躍客戶，任何時間點都有數十個專案在運行。

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.

這些結果現在終於都回傳到 LensAI 入口網站中。因此，這些團隊會收到安全登入資訊；登入後即可存取這個入口網站，並看到其中的應用程式，而這些應用程式確實改變了他們過去進行藥物發現的方式。

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 for it and worked for it. 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.

現在他們可以使用這些應用程式；不必再去找 2、3、4、5、6 家其他供應商蒐集資訊，或在 18 個月到 2 年的時間裡一路「硬撐」著把流程走完；他們可以直接以訂閱方式，在基礎層級之上使用這些應用程式，發揮其效能並取得他們想要的結果。因此，能走到這個階段對我們公司而言非常重要。這是我們為之打造、也為之努力的成果。把這套軟體交到這些客戶手上所花的時間比我們原先希望的更久，但現在正在發生，而且不僅涵蓋我們的治療領域客戶，也涵蓋那些與我們簽約、實際上是為了任何形式客製化抗體工作的客戶。

And so with regard to that, when you 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.

因此，談到新增合約與導入新客戶——這正是我們真正聚焦的地方——我們認為我們處於一個極其獨特的情境：這些客戶其實已經完成導入（onboarded）。在許多情況下，我們是主要供應商；但在所有情況下，我們都是已經在他們供應商系統中的供應商。而且我們已經建立了他們的信任與尊重。因此，透過這些客戶，我們得以以非常獨特的方式順勢切入這個市場。

Thanks for that detailed. 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 under the influence 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?

謝謝你這麼詳細的說明。如果可以的話，我還有第二個問題，關於資產層級融資（asset level financing）。我確實理解律師和投資人往往需要很長時間才能對任何事情達成結論，但就你們正在考量的這 4 個不同專案／平台而言——登革熱、GLP-1、Llama，以及其他這些（你說的第四個是 under the influence）——你們需要在多大程度上等這些團體達成結論，才會把這些專案往前推進？還是說它們彼此獨立、都在往前推進？

So that is a great question. So the short answer is they're independent of 1 another as they move forward. A couple of things to keep in mind. When we look at financing these particular programs I think that 1 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 is conducted in-house.

這是個很好的問題。簡短的答案是：它們在推進上彼此獨立。有幾點需要記住。當我們評估為這些特定專案融資時，有一點很容易被忽略：在這個階段，我們的專案成本並不像傳統藥物開發公司那樣高。因此，我們大量工作是在電腦模擬（in silico）中完成；而且我們的 in silico、in vitro 甚至臨床前工作，很多不是由 AI 驅動，就是在內部完成。

So that keeps our cost meaningfully lower than the conventional pipeline of this breadth would require. And it's also one of the structural advantages that we have building on the HYFT platform.

因此，這讓我們的成本顯著低於同等廣度的傳統研發管線所需的成本。這也是我們在 HYFT 平台基礎上建構所具備的結構性優勢之一。

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 decel 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, 1 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 such a better experience, which with each of these families of viruses or the particular therapeutic or disease that we're targeting.

因此，基於這一點，並直接回應你的問題，RK，我們目前擁有的資金足以在這些合作推進中把我們帶得相當往前。事實上，Scott 提到的一點是研發費用，與去年相比並沒有大幅增加，但卻不僅涵蓋我們傳統的研發，以及 decel 平台的建置擴張，也涵蓋了我們迄今在這裡所做的一切。我想這算是一個相對具體的例子。至於資產層級融資，這確實是當這些專案更進一步成熟時會需要的；另外有一點我們沒有談太多，但我們在推進過程中已確保到位的是：我們有一支專業團隊，具備臨床領域的經驗，以及針對我們鎖定的各病毒家族、或特定治療領域／疾病的更深厚經驗。

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.

以協助推動流程，完成臨床前階段、IND 使能（IND-enabling）、IND 申報，以及臨床就緒。當我們進入那些階段時，成本當然就會開始上升。至於在那個階段是否能在資產層級融資之前繼續推進，在某種程度上是可以的——而且絕對可以——再次強調，因為我們已經在這裡配置了具備內部專業能力的團隊。但除此之外，資產層級融資的目的，是在我們進入那個階段後提供支持。我們目前的資金跑道足夠，因此實際完成這些專案的隔離／圈定（ring-fenced）所需的前置時間，應能讓我們在那之前引入額外資金，以支持屆時的需求。

Thank you. I will step back into the queue. Thanks.

謝謝。我先回到隊列裡。謝謝。

Thanks, RK.

謝謝你，RK。

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

目前沒有其他問題。我現在把電話會議交回給 Bath 博士作結語。

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.

很好。非常感謝。好。我最想說的一件事就是：謝謝大家。謝謝各位加入我們。

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.

謝謝各位支持 MindWalk。我們期待在管線結果出爐後與各位分享。我們將在 2026 會計年度第四季與年度結束的 '26 財報電話會議上與各位再度交流。謝謝。

This concludes today's MindWalk Holdings Q3 fiscal year 2026 earnings call. Thank you for your participation. You may now disconnect.

以上為 MindWalk Holdings 2026 會計年度第三季財報電話會議。感謝各位參與。您現在可以掛線。