Dognosis Pairs Rescue Dogs With AI to Screen Cancer From Home Masks

Fireside Chat with Amir Alpert, Dognosis — Ai4

Ai4

By digitizing canine olfactory brain responses to volatile organic compounds on mail-in breath masks, Dognosis aims to bypass healthcare bureaucracy and make multi-cancer screening as cheap as a streaming subscription.

AI-assisted summary of Fireside Chat with Amir Alpert, Dognosis — original source. Check the original for full context.

Source published: 2026-08-17

Detailed notes

## Background Published on YouTube by Ai4 on August 17, 2026, this video features a fireside chat with Amir Alpert, Head of US Business Development at Dognosis. The session covers the startup's approach to merging canine olfactory biology with machine learning to build non-invasive diagnostic tools and advance interspecies communication.

## TL;DR Amir Alpert presents Dognosis alongside Snow, a rescue dog, detailing how the company combines canine scent detection and automated AI analytics to screen for early-stage cancer using simple mail-in breath masks. The company aims to make routine cancer triage as accessible and affordable as a monthly streaming subscription while gathering data for future insurance coverage. Beyond oncology, Dognosis utilizes non-invasive EEG brain-computer interfaces through its 'DogSense' platform to pioneer interspecies communication and build olfactory AI models. Alpert also emphasizes an ethical operational model that relies on rescued dogs and community-driven AI development rather than commercial breeding programs.

## Atmosphere & style The presentation is warm, candid, and visionary, effortlessly balancing complex biotech and machine learning concepts with genuine affection for canine companions. Amir brings an earnest, conversational enthusiasm to the stage, speaking with passionate conviction about animal welfare and healthcare accessibility. The room feels energized and charmed by Snow's presence, creating an uplifting, emotionally resonant dialogue that culminates in a communal, celebratory crowd selfie.

## Key takeaways - [00:04](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=4s) Dognosis announced the launch of its US headquarters at AI4, showcasing how canine biological olfaction combined with machine learning can address major global health challenges. - [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) The company's core oncology mission echoes Reed Jobs' vision: transforming cancer detection and treatment into an experience as manageable and non-lethal as the seasonal flu. - [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) Beyond oncology, Dognosis's broader technological North Star is establishing two-way interspecies communication to decipher animal cognition and physical states. - [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) Amir Alpert argued against the tech industry's narrow focus on balance-sheet expansion, contending that AI's true measure of value is improving the daily lives of living beings. - [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) Dognosis detects early-stage cancer by capturing volatile organic compounds (VOCs) exhaled into a mail-in N95 breath mask worn at home for just ten minutes. - [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) To achieve reproducible FDA-level diagnostic standards, the testing facility pairs trained canine scent stations with AI algorithms and video recording to eliminate human handler bias. - [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) The company's tech roadmap outlines three 'superpowers': Multi-Cancer Early Detection, a Smell Positioning System (SPS) scent GPS, and Olfactory AI that teaches computers how to smell. - [08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) Dognosis developed 'DogSense', a non-invasive canine Brain-Computer Interface (BCI) adapting EEG technology to translate olfactory detection into objective neural spike data. - [08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) Dognosis emphasizes a strict ethical commitment to animal welfare, positioning rescue dogs as respected biological partners rather than traditional research test subjects. - [11:10](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=670s) The company plans to price its consumer screening kits affordably, comparing the intended cost of regular cancer monitoring to an average household streaming subscription. - [11:10](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=670s) To circumvent bureaucratic delays in healthcare reimbursement, Dognosis launched via a direct-to-consumer pathway while accumulating clinical evidence for future payer coverage. - [13:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=834s) High-performing scent-detection capability is innate across breeds; Dognosis relies primarily on shelter rescues and Indian street dogs rather than expensive pedigrees. - [13:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=834s) Co-founder Ákos Kollár leveraged generative AI over a single weekend to 'vibe code' an internal application for managing the health, enrichment, and tracking of their rescue pack.

## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=4s) Introducing Dognosis and Defining "AI for Good"

Host Lurin kicks off the session at AI4 by acknowledging that although a prior session had been canceled, this particular fireside chat was his personal favorite to program across the entire conference. The appeal extends far beyond having a charming four-legged guest on stage; it represents a tangible technological breakthrough designed to address global human health challenges. Lurin introduces Amir Alpert, Head of US Business Development at Dognosis, alongside the undisputed star of the event: Snow, a white rescue dog attending his first-ever tech conference. Amir gives a playful introductory voice to Snow, greeting the room with, "I'm Snow. I'm the first dog in Dognosis. Any questions, just ask my dad. I'm not yet able to talk back, but I hear our company is working on that dog speed."

When asked to unpack the often vague industry catchphrase "AI for Good," Amir explains that attendees spend entire conferences surrounded by dazzling algorithms while wondering how the technology will actually solve humanity's deepest crises. Dognosis chose the AI4 platform to announce the official launch of its US headquarters to demonstrate how cutting-edge machine learning and biological intelligence can coalesce into life-saving medical applications.

[02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) The North Star: Taming Cancer and Interspecies Communication

Amir outlines the ultimate vision driving Dognosis. Rather than claiming they will eradicate or "cure" cancer so that no human ever develops it, Amir believes Snow and the canine team represent a pivotal puzzle piece in ensuring cancer no longer devastates human lives. Echoing Reed Jobs, the founder of the oncology venture fund Yosemite, Amir emphasizes that detecting and treating cancer should eventually become as manageable as having the flu.

Beyond oncology, the company’s ultimate technological North Star is interspecies communication. Amir envisions expanding beyond dogs to household pets, racehorses, and cats, using AI to replace human guesswork with educated, real-time insight into animal cognition and physiological states. This transition elevates animals from simple domestic pets into recognized biological partners. Amir challenges the prevailing tech ethos championed by enterprise giants and major chipmakers like Nvidia, noting that founders and investors fixate too heavily on balance-sheet expansion. He argues that true AI value lies in elevating the day-to-day lives of living beings:

> "Focus on how AI improves the lives of the people that use it and of the lives that they can have when they're not using it... Once we accept that the mission here is to improve the lives of everyone that interacts with AI, then we will see the value and the pie will be so big, we will sit in rooms like this and go, why did we ever question the money?"

- Reed Jobs' benchmark: transforming a cancer diagnosis into an experience no more lethal or disruptive than contracting the seasonal flu. - Company scale: an early-stage startup funded by a seed round of a few million dollars, aiming to shift priorities across big tech. - Long-term horizon: translating canine sensory signals into actionable data across veterinary medicine, sports, and companion animal welfare.

[05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) Breath Masks, Volatile Biomarkers, and the Three Superpowers

Transitioning into the operational mechanics of Dognosis, Amir breaks down how canine olfactory biology is converted into actionable diagnostic data. Cancer does not emerge in an instantaneous flash; it represents an extended progression of metabolic disruptions spanning years. As tissues experience metabolic shifts, distinct volatile organic compounds (VOCs) are released into the bloodstream, transferred across the alveoli into the lungs, and exhaled in the breath. Rather than subjecting patients to high-stress, delayed hospital visits, Dognosis utilizes a simple, non-invasive home screening kit.

```mermaid flowchart TD accTitle: Dognosis Breath Testing Workflow accDescr: Diagram illustrating the end-to-end process from patient breath collection to AI-verified canine diagnostic scoring.

subgraph Collection["Patient Sampling"] A["Patient breathes into N95 mask<br/>(10 mins at home)"] --> B["Mail-in test kit sent to Dognosis lab"] end

subgraph Testing["Automated Canine & AI Lab"] B --> C["Trained canine pack inspects VOC scent lines"] C --> D["Automated sensory stations & cameras record response"] end

subgraph Analytics["Diagnostic Output"] D --> E["AI algorithms evaluate neural & behavioral data"] E --> F["Standardized Canine Reactivity Score generated"] end ```

The testing protocol eliminates standard medical friction: a patient wears a standard COVID-style N95 breath mask at home for ten minutes while reading the morning paper or solving a Wordle puzzle, then mails the kit back. At the facility, automated testing rigs present the breath sample to a trained pack of dogs. High-definition recording systems and quality-controlled algorithms monitor the canine inspection, removing human observer bias and creating an automated, reproducible platform capable of meeting FDA-level diagnostic standards.

Amir categorizes Dognosis's broader technological roadmap into three core superpowers: - **Superpower 1: Multi-Cancer Early Detection & Triage:** Utilizing canine scent sensitivity to catch malignant changes long before conventional symptoms appear, functioning as an entry triage tool that can later expand to Parkinson's, Alzheimer's, and diverse blood disorders. - **Superpower 2: Smell Positioning System (SPS):** Translating the invisible olfactory landscape into an intuitive visual map for humans, operating as a sensory GPS based on environmental scent distribution. - **Superpower 3: Olfactory AI:** Reverse-engineering canine sensory processing to develop artificial intelligence capable of smelling, adding a third fundamental sensory dimension alongside computer vision and natural language processing.

[08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) Ethical Animal Partnership and the DogSense BCI

Amir confronts the sensitive history of animal use in scientific research, directly addressing common skepticism seen on social media regarding canine welfare. Speaking candidly about Snow, who has lived as his companion for seven years, Amir makes an impassioned personal commitment:

> "Will God be my witness? This dog has been with me through thick and thin. If I signed him up, I guarantee you we are treating these dogs with humanity that will inspire humans to be better humans to other humans."

To substantiate how animal communication is digitized, Amir presents an actual scan of Snow's brain activity collected via "DogSense"—a non-invasive brain-computer interface (BCI) adapting decades of validated human electroencephalogram (EEG) technology to canine physiology. When a dog identifies a target scent—a principle long utilized by military and border patrol units for explosives and narcotics—their brain produces measurable electrical spikes. Demonstrating an EEG readout on screen, Amir displays sharp electrical peaks triggered when Snow caught the scent of salami, proving that canine interest can be tracked as an objective, digitized neurological event rather than relying on subjective tail wags or handler interpretations.

[11:10](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=670s) Democratizing Diagnostics: Streaming Subscriptions and Insurance

Addressing the commercialization and scalability of Dognosis in the United States, Lurin asks how the company plans to navigate consumer adoption and the notoriously bureaucratic American healthcare reimbursement landscape. Amir turns to an audience member named Sash to compare costs with modern digital media habits. While Sash estimates his household spends roughly $20 to $30 a month on entertainment, average American households spend approximately $77 monthly on streaming services.

Amir reveals that Dognosis intends to price its consumer screening kit so accessibly that families will view regular cancer monitoring with the same financial nonchalance as maintaining their streaming subscriptions. While formal insurance coverage remains a vital pillar for widespread clinical adoption, gaining payer reimbursement requires extensive longitudinal trial data and regulatory sign-offs. To avoid years of institutional delays, Dognosis has structured a legally vetted direct-to-consumer pathway, allowing proactive consumers to monitor their health immediately while the company gathers the clinical data needed for future insurer coverage.

[13:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=834s) Dog Sourcing, Weekend "Vibe Coding", and the Community Selfie

Lurin questions whether deploying scent-detection dogs requires elite, costly bloodlines similar to thoroughbred racing horses. Amir dispels the notion of genetic elitism, explaining that scent capability is broadly innate across canines. Snow himself was adopted from Houston in the aftermath of Hurricane Harvey seven years ago. At Dognosis's primary facility in Bangalore, India, the team manages a pack of 15 dogs; while some are beagles—a breed traditionally prized in scent detection—many others are rescued Indian street dogs.

Amir highlights co-founder Ákos Kollár, who used modern generative AI tools over a single weekend to "vibe code" a dedicated application that coordinates communal tracking, enrichment, and medical care for their rescue pack. This low-cost, high-compassion operational model proves that scaling does not require breeding hundred-dog litters only to discard the vast majority.

Before closing the session, Amir pauses to prevent early exits, inviting the entire room to participate in a signature tradition: the Dognosis crowd selfie, celebrating Snow's very first flight to a tech conference. Amir displays a QR code on screen and asks the audience to amplify their mission:

> "We don’t want your money. We want your awareness. We want you to tell your friends. We want you to get inspired. We want you to sit around coffee and say, 'Did you hear about the company that’s using dogs? Did you hear about the company that wants to give us the tools to talk to dogs?' You give us that and we are forever in your service."

## Notable quotes > [00:04](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=4s) I'm Snow. I'm the first dog in Dognosis. Any questions, just ask my dad. I'm not yet able to talk back, but I hear our company is working on that dog speed. — Amir Alpert

> [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) Focus on how AI improves the lives of the people that use it and of the lives that they can have when they're not using it... — Amir Alpert

> [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) Once we accept that the mission here is to improve the lives of everyone that interacts with AI, then we will see the value and the pie will be so big, we will sit in rooms like this and go, why did we ever question the money? — Amir Alpert

> [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) Multi-Cancer Early Detection & Triage — Amir Alpert

> [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) Smell Positioning System (SPS) — Amir Alpert

> [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) Olfactory AI — Amir Alpert

> [08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) Will God be my witness? This dog has been with me through thick and thin. If I signed him up, I guarantee you we are treating these dogs with humanity that will inspire humans to be better humans to other humans. — Amir Alpert

> [08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) DogSense — Amir Alpert

> [13:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=834s) We don’t want your money. We want your awareness. We want you to tell your friends. We want you to get inspired. — Amir Alpert

> [13:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=834s) Did you hear about the company that’s using dogs? Did you hear about the company that wants to give us the tools to talk to dogs? — Amir Alpert

> [13:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=834s) You give us that and we are forever in your service. — Amir Alpert

## Mentioned Dognosis, Amir Alpert, Lurin, Ai4, Snow, Reed Jobs, Yosemite, Nvidia, DogSense, Ákos Kollár, Hurricane Harvey, Wordle, FDA

## Video description Amir Alpert - Head of US Business Development, Dognosis

[Source](https://www.youtube.com/watch?v=DRKW9zjtxRs)

中文简报

Dognosis结合救助犬与AI，通过居家口罩筛查癌症通过将犬类对邮寄呼吸口罩上挥发性有机化合物的嗅觉大脑反应数字化，Dognosis旨在绕过医疗官僚流程，让多癌种筛查像流媒体订阅一样实惠。 Amir Alpert与救助犬Snow一同展示了Dognosis项目，详细介绍了该公司如何将犬类嗅觉检测与自动化AI分析相结合，利用简便的邮寄呼吸口罩来筛查早期癌症。该公司的目标是让常规癌症分流筛查像每月流媒体订阅一样易于获取且价格亲民，同时为未来纳入医保收集数据。除肿瘤筛查外，Dognosis还通过其“DogSense”平台利用非侵入性EEG脑机接口，探索跨物种交流并构建嗅觉AI模型。Alpert还强调了一种合乎伦理的运营模式，即依靠收容所救助犬和社区驱动的AI开发，而非商业繁育计划。 Dognosis 在 AI4 大会上宣布设立美国总部，展示结合犬类生物嗅觉与机器学习来应对全球重大健康挑战的技术。 (0:04) 公司的核心肿瘤愿景呼应 Reed Jobs 的理念，致力于将癌症检测与治疗转变为像季节性流感一样可控且不致命的日常体验，并拓展跨物种交流的远景。 (2:06) Dognosis 采用非侵入式居家检测套件，患者佩戴 N95 口罩 10 分钟后寄回，由自动化犬类气味台配合 AI 与视频记录消除人为偏差，以达到 FDA 级别的筛查标准。 (5:00) 团队公布了三大技术能力：多癌症早期筛查、气味定位系统（SPS）以及嗅觉 AI，并开发出名为 DogSense 的非侵入式脑机接口（BCI），将犬类嗅觉反应直接转化为客观的神经电信号数据。 (8:54)

Reading summary

Dognosis combines canine olfaction with non-invasive brain-computer interfaces and machine learning to develop low-cost, early-stage cancer screening and interspecies communication tools, according to Amir Alpert, Head of US Business Development. Early Cancer Triage and Olfactory AI (5:00) Dognosis aims to transform cancer detection into a manageable health event comparable to the flu, referencing a benchmark articulated by Yosemite founder Reed Jobs. Instead of relying immediately on invasive hospital visits, the company's screening model targets the volatile organic compounds (VOCs) that metabolic cellular changes release into the bloodstream and breath over years. The process uses a home kit where a patient wears an N95 mask for ten minutes, which is then mailed to a testing facility. At the testing site, a pack of trained dogs evaluates the scent samples within automated rigs recorded by cameras, feeding behavioral and neurological data into machine-learning algorithms to produce a standardized diagnostic score. Alpert outlines a three-part technical roadmap: multi-cancer early detection and triage (with future aspirations for Parkinson's and Alzheimer's), a "Smell Positioning System" (SPS) to map environmental scents visually, and synthetic "Olfactory AI" that reverse-engineers canine sensory processing to create machine-based scent detection. Digitizing Canine Signals via DogSense BCI (8:54) To counter ethical concerns about using animals in scientific testing and eliminate handler subjectivity, Dognosis utilizes "DogSense," a non-invasive canine brain-computer interface adapted from human electroencephalogram (EEG) technology. By measuring real-time electrical neural activity, the system turns scent recognition into an objective digital event rather than relying on ambiguous physical signals like tail wagging. Demonstrating this with Snow—a white rescue dog adopted after Hurricane Harvey—Alpert showed an EEG trace displaying sharp electrical peaks recorded when the dog smelled salami. Alpert contends that objective neural telemetry provides an ethical framework that treats working canines as biological partners, aiming to extend this communication technology to other animals such as cats and racehorses. Direct-to-Consumer Model and Rescued Packs (11:10) Navigating the slow adoption cycle and regulatory hurdles of the American healthcare reimbursement landscape, Dognosis plans an initial direct-to-consumer subscription model. Alpert estimates this out-of-pocket screening will cost roughly on par with household digital streaming services, which average about $77 monthly in the United States, allowing the company to build longitudinal data while pursuing long-term insurance reimbursement and FDA-level standards. The operational model does not rely on purpose-bred canine bloodlines. At the company's Bangalore facility, the 15-dog pack consists of beagles alongside rescued Indian street dogs. Alpert notes that co-founder Ákos Kollár developed a communal tracking and enrichment app over a weekend using generative AI tools to manage health and operational workflows for their rescue pack. Dognosis remains an early-stage startup supported by a seed round of a few million dollars, facing the task of validating its canine-AI diagnostic platform to secure eventual FDA recognition and formal healthcare payer reimbursement.

精炼阅读摘要

Dognosis 美国业务发展主管 Amir Alpert 提出，将犬类嗅觉生物学特性与非侵入性脑机接口及人工智能相结合，能够构建高可及性、无创的泛癌症早期筛查机制。该团队主张将 AI 的核心目标从单纯的商业算力扩张转向改善生命福祉，旨在使癌症早筛与干预像应对流感一样可控，并最终建立以嗅觉数据为核心的跨物种沟通体系。重新定义医疗愿景与跨物种沟通 (2:06) Amir Alpert 在 AI4 大会上阐述了 Dognosis 的核心理念，即通过结合机器学习与生物智能来应对人类重大健康危机。他引用肿瘤风险投资基金 Yosemite 创始人 Reed Jobs 的观点指出，公司并不宣称能够彻底消除人类罹患癌症的可能，而是致力于在极早期介入诊断，使癌症的发现与治疗体验转变为像季节性流感一样不再造成严重生活破坏与高致死率。Alpert 同时批评了以芯片制造巨头 Nvidia 为代表的科技巨头过分聚焦于资产负债表与财务增长的行业风气，强调人工智能的真正价值在于切实改善生命福利。除肿瘤医学外，该团队还将跨物种沟通确立为长远的技术指引。Alpert 表示，Dognosis 的目标是逐步从犬类拓展至赛马、猫等家庭与伴侣动物，借助 AI 实时洞悉动物的生理状态与认知活动，使动物从传统的家养宠物转变为经科技确认的生物学合作伙伴。这家目前仅获得数百万美元种子轮融资的早期初创企业，试图推动整个科技行业反思算力堆叠背后的终极价值取向。呼气代谢物检测与三大嗅觉技术路径 (5:00) Dognosis 的检测机制基于癌症在多年演进过程中产生的代谢紊乱。组织代谢发生异常后，特定的挥发性有机化合物（VOC）会进入血液循环，经肺泡扩散至肺部并通过呼吸排出。公司设计了一套无创的居家筛查方案：受检者只需在家佩戴标准 N95 呼气口罩 10 分钟，即可邮寄至其实验室。在检测设施内，自动化测试台将样本呈送给受训犬群，高清晰度记录设备与算法全程追踪犬类的嗅探行为，生成标准化的犬类反应评分，从而剔除人工观察者偏见，并以此寻求达到 FDA 级别的医疗诊断标准。 Alpert 将 Dognosis 的技术演进路线划分为三大超能力：首先是泛癌症早期筛查与分诊，利用犬类嗅觉灵敏度在常规症状出现前捕获恶性病变，未来规划拓展至帕金森病、阿尔茨海默病及血液系统疾病；其次是气味定位系统（SPS），将肉眼不可见的气味分布转化为直观的视觉地图，实现环境气味导航；最后是嗅觉 AI（Olfactory AI），通过逆向工程解析犬类感觉处理机制，赋予人工智能嗅觉维度的感知能力，与现有的计算机视觉和自然语言处理并列。脑机接口与客观神经信号解码 (8:54) 针对科研实验中涉及动物福利的历史遗留问题和公众质疑，Alpert 强调了团队的人道原则，并以陪伴自身七年的救援犬 Snow 为例保证动物在研究过程中的待遇。为了从根本上消除依赖犬只摆尾或驯犬员主观解读带来的诊断误差，Dognosis 开发了名为“DogSense”的非侵入性脑机接口（BCI）。该技术借鉴了人类脑电图（EEG）数十年的成熟原理并将其适配到犬类生理结构上。在会议现场的演示中，Alpert 展示了 Snow 的实时脑电活动图谱。当犬只辨别出目标气味（例如演示中嗅探到萨拉米香肠）时，其大脑会产生显著且可测量的电信号峰值。这表明犬类的嗅觉识别行为可以被记录为客观、精确量化的数字化神经事件，从而为大规模自动化诊断提供了不受人为干扰的生物神经数据验证支撑。消费级订阅定价与低成本运维 (11:10) 面对美国医疗健康体系繁琐且漫长的医保报销审批，Dognosis 采取了直接面向消费者（D2C）的市场落地路径。Alpert 将检测套件的定价参照现代数字媒体订阅习惯，指出普通美国家庭每月在流媒体服务上的支出约为 77 美元，而 Dognosis 计划将其筛查产品定价降低至让家庭能够以订阅流媒体的心态无压力负担。通过 D2C 渠道，企业可以在规避体制延误的同时立即提供监测服务，并同步累积支撑后续申请正式医疗保险报销所需的长期临床数据。在犬只选拔与运营层面，Alpert 破除了必须依赖昂贵名贵血统犬的传统观念，指出气味辨识能力是犬类的广泛天性。在其位于印度班加罗尔的基地中，15 只受训犬不仅包含嗅觉能力突出的米格鲁猎犬，也包括多只印度流浪救援犬，Snow 本身也是七年前在飓风哈维后获救的搜救犬。联合创始人 Ákos Kollár 还利用现代生成式 AI 工具在周末快速“氛围编程”（vibe coding）出犬群管理应用，以低成本数字工具协同犬群的日常照护、健康追踪与医疗保障，证明了该体系无需依赖大规模工业化繁殖即可实现扩容。 Dognosis 试图通过融合救助犬生物嗅觉、脑机接口与算法分析，开辟一条低门槛、无创的癌症早筛与嗅觉计算途径。不过，其商业与临床前景仍取决于能否在规避医保报销壁垒的 D2C 阶段积累充足的纵向试验数据，以最终获得 FDA 及主流医疗支付方的正式认可；目前公司正优先通过社区传播拓展公众认知，而非单纯寻求资本扩张。

Detailed notes — zh

## Background 该视频由 YouTube 频道 Ai4 于 2026 年 8 月 17 日发布。根据页面简介，内容为 Dognosis 美国业务发展负责人 Amir Alpert 在大会现场展开的炉边对话访谈。 ## TL;DR 本场对话围绕生物科技初创公司 Dognosis 的创新探索展开，介绍了其如何将犬类极其敏锐的嗅觉与 AI、脑机接口（DogSense）相结合，通过仅需佩戴 10 分钟口罩的呼气采样，实现低负担、非侵入性的多癌种早期筛查。团队的终极愿景是消除对癌症的致命恐惧并实现跨物种沟通，让严重疾病的早期干预像应对流感一样从容。在落地策略上，Dognosis 绕开传统医保繁冗的早期审批，以对标流媒体订阅的平民化 D2C 模式推向大众，并倡导善待救援犬与流浪犬，以真正的“AI 向善”守护生命健康。 ## Atmosphere & style 现场氛围既轻松幽默又充满温情与使命感。救援犬 Snow 的同台亮相为严肃的硬核前沿技术对话增添了生动可爱的气息，主讲人 Amir 借爱犬之口与台下幽默互动，迅速打破了传统科技演讲的刻板距离。随着话题从呼气早筛深入到犬类脑电信号、动物福祉以及对科技巨头执迷于财务报表的犀利反思，整个对话的情绪逐渐升华为对生命价值的深刻关怀，最后在全场观众温暖热烈的自拍合影与认知倡导中达到高潮。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=4s) Dognosis 提出将 AI 与犬类嗅觉感知结合，致力于将其打造为践行“AI 向善”（AI for Good）的硬核典范。 - [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) 公司的长期愿景是降低癌症致命伤害，实现像应对流感般的早期筛查干预，并进一步探索跨物种沟通。 - [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) 呼气多癌种早筛流程极为便捷，用户仅需在家佩戴专用 N95 口罩 10 分钟即可完成样本采集。 - [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) 通过自研 AI 自动化采样与算法质控平台，Dognosis 能够客观量化犬类的嗅觉神经与行为反应，消除人工判断偏差。 - [08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) Dognosis 研发了基于脑电图（EEG）技术的 DogSense 犬类脑机接口，可直接读取并可视化工作犬识别特定气味时的神经兴奋信号。 - [10:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=654s) 商业定价目标是将癌症早筛门槛降低至对标流媒体订阅费用的平民化水平，减轻大众的财务与心理负担。 - [12:35](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=755s) 在市场准入方面，团队率先采取直接面向消费者（D2C）的合规路径，并在积累充分临床数据后逐步融入传统医保报销体系。 - [13:56](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=836s) 检测犬选拔门槛极具包容性，涵盖飓风救援犬、小猎犬及街头流浪犬，绝大多数普通犬类经训练均具备出色的嗅觉识别潜力。 - [15:17](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=917s) 团队呼吁公众提升对犬类嗅觉检测与跨物种沟通技术的认知度并主动传播，以加速该技术造福人类健康。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=4s) 开场介绍与 Dognosis 的“AI 向善”理念在 AI4 大会的现场，主持人 Lurin 热情地开启了这场炉边对话。尽管前一场活动遗憾取消，但他直言这场对话是整个大会中他最期待的环节。除了现场有一只可爱的狗狗吸引眼球之外，更重要的是其背后承载着有望改变世界的硬核技术。在正式展开讨论前，现场进行了趣味介绍：Dognosis 的美国业务发展负责人 Amir Alpert 带着他的爱犬 Snow 一同登台。Amir 借着 Snow 的口吻幽默地向大家打招呼：“我是 Snow，Dognosis 的第一只狗狗。有任何问题尽管问我爸爸。虽然我现在还不会开口说话，但我听说公司正在紧锣密鼓地研发这项跨物种技术。” 现场气氛在笑声中迅速活跃起来。主持人随即切入核心议题：在 AI4 充斥着各种企业级 AI 应用的背景下，Dognosis 的技术究竟如何诠释“AI 向善”（AI for Good）这一概念？一只狗与人工智能之间看似毫无关联，二者又是如何产生交集的？Amir 表示，会场里的每位从业者都在见证惊人的技术爆发，也都在思考 AI 将如何解决世界上最重大的挑战。Dognosis 正式在此设立美国总部，正是为了向身处前沿的顶尖人才展示一个清晰的范例——如何真正将 AI 应用于造福人类福祉的伟大事业中。 --- ### [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) 终极目标：让人类像对待流感一样面对癌症与跨物种沟通谈及公司的“北极星指标”（North Star）与长期愿景，Amir 分享了他们在癌症筛查领域的突破性探索。他明确表示，自己不会轻率地使用“彻底治愈癌症”这种字眼，但这支由狗狗与 AI 组成的团队确实是攻克癌症难题的关键拼图。他引用了癌症医疗风投基金 Yosemite 创始人 Reed Jobs 的名言——未来的目标是让患上癌症“就像得了一场流感一样”，不再给人类带来昔日那种毁灭性的伤痛与恐惧。除了癌症筛查，Dognosis 更宏大的远景是实现“跨物种沟通”（Interspecies Communication）。从家养的猫狗宠物，到赛场上的骏马，AI 将帮助人类打破物种壁垒，借助数据洞察和智能引导去真正读懂动物的所思所感，使动物从被动驯养的“宠物”升格为平等的“合作伙伴”。针对当下整个科技界对 AI 价值的焦虑，Amir 直言不讳地指出，行业过度受限于“资产负债表是否变大”的金钱视角。他公开向科技巨头呼吁，应当把注意力从单纯的商业变现转向如何实质性改善生命质量：“这才是真正的 AI 向善。一旦我们确立了这个使命，创造出的整体价值蛋糕将无可限量，到那时我们甚至会反思，当初为何要对金钱如此狭隘地执迷。” - **核心愿景**：降低癌症致命伤害，实现类似流感般的早期干预；打通跨物种沟通管道。 - **价值理念**：超越财务报表驱动的传统科技视角，以改善人类与生命福祉为首要衡量标准。 --- ### [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) 核心技术与三大超能力：呼气早筛与嗅觉人工智能 Amir 结合幻灯片深入浅出地拆解了 Dognosis 所依托的核心技术。他们将犬类的生物学天赋转化为三大“超能力”：第一项超能力是“无负担的多癌种早期筛查”（Multicancer Early Detection）。传统的体检往往充斥着令人抗拒的就医流程，而 Dognosis 的方案极其轻便：用户只需在家佩戴标准的 N95 呼吸口罩约 10 分钟，在此期间可以悠闲地看报纸或玩填字游戏，随后将样本寄回实验室。 ```mermaid flowchart TD accTitle: Dognosis 呼气癌症筛查与自动化分析流程 subgraph S1["用户端采集"] A["居家佩戴专用口罩 10 分钟"] --> B["样本吸附呼气代谢分子并寄回"] end subgraph S2["犬类嗅觉感知"] B --> C["受训犬群嗅闻检测样本"] C --> D["捕捉犬类特异性神经与行为反应"] end subgraph S3["自动化与算法质控"] D --> E["AI 自动化测试与记录平台"] E --> F["输出犬类反应评分 (K9 Reactivity Score)"] F --> G["生成疾病分诊与早期预警建议"] end ``` 人体癌症并非瞬间爆发，而是一个伴随代谢异常的长期病理演进过程。受损组织释放出的特定代谢分子会从血液循环进入肺部，并最终随呼气排出，而犬类拥有极为敏锐的嗅觉感知系统。Dognosis 的关键突破在于“闭合人犬沟通回路”——研发出自动化的训练、采样、测试与记录平台，彻底消除人工判定带来的主观偏差，以对标 FDA 药物级审批的严格质控算法来标准化犬类的筛查结果。这一模式未来还将拓展至阿尔茨海默病、帕金森病及血液系统疾病的分诊。此外，公司的第二项超能力是构建可视化的“嗅觉定位系统”（Smell Positioning System, SPS）；第三项超能力则是对嗅觉机理进行逆向工程，开发真正“具备嗅觉感知能力的 AI 硬件与底层架构”。 - **筛查原理**：利用犬类嗅觉捕捉呼气中由早期癌变与代谢失调释放的挥发性生物标志物。 - **流程体验**：10 分钟居家呼气采样，摆脱传统医院繁琐预约与侵入性检查。 - **技术突破**：打造标准化 AI 自动化采样与算法质控平台，实现诊断级客观评估。 - **拓展领域**：从癌症逐步延伸至阿尔茨海默病、帕金森病等各类重大疾病的早期分诊。 --- ### [08:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=534s) 动物福祉与 DogSense 脑机接口：解读犬类大脑信号谈及实验动物的待遇问题，Amir 展现出了真诚而严肃的态度。他坦言，历史上科学界对待动物的方式确实存在过失误，社交媒体上也常有公众担忧工作犬的生存处境。对此他明确担保，Snow 已经陪伴他度过了七年的风雨岁月，如果让自己的爱犬参与其中，必然会以最高标准的人道关怀对待所有犬只，绝不辜负公众的信任。随后，Amir 展示了团队自主研发的犬类脑机接口系统——DogSense。该设备将人类医学领域数十年成熟验证的脑电图（EEG）技术创新性地应用于犬类。当工作犬佩戴该设备时，大脑皮层的电生理刺激会被实时采集并可视化。正如军用排雷犬或缉毒犬在发现目标时表现出的兴奋状态一样，医疗犬在嗅到癌症或特定气味标志物时也会在大脑中产生特征性电信号。为了让观众直观理解，现场大屏幕展示了 Snow 佩戴 DogSense 嗅闻意大利萨拉米香肠时的实时脑电波形图，屏幕上清晰显现的一个个脉冲峰值，精准记录下了犬类大脑在识别出特定气味那一瞬间的强烈神经兴奋。 - **DogSense 脑机接口**：基于成熟 EEG 脑电技术，实时采集并解码犬类嗅觉神经信号。 - **客观量化**：摆脱单纯肉眼观察犬类肢体动作的传统模式，直接读取神经电生理反馈。 - **伦理保障**：坚持人道主义关怀标准，确保科研犬只在健康、友好的环境中工作。 --- ### [10:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=654s) 商业落地构想：对标流媒体订阅的平民化癌症筛查当话题转向商业化落地与复杂的美国医疗保险体系接入时，Amir 抛出了一个生动的对比。他向现场来自班加罗尔等地的参会观众提问，询问大家每月在各类流媒体平台上的总支出是多少。观众回答大约为 20 至 30 美元，而 Amir 指出全美家庭在流媒体上的平均月支出高达 77 美元。 Dognosis 的商业定价策略正是围绕这一心理账户展开：尽管目前在美国尚处于早期阶段，定价尚未最终敲定，但团队的目标是将防癌筛查的费用降低到“如同订阅流媒体服务一样无需犹豫”的平民化水平。Amir 总结道，绝大多数人都清楚定期筛查的重要性，但现实生活的繁琐往往让人顾此失彼。在 AI 技术带来的充裕时代（Abundance Mindset），筛查技术不应再被高昂的门槛与复杂流程所束缚，而应当以最低的心理与财务负担，真正守护每个普通人的生命健康。 ### [12:35](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=755s) 医疗保险体系的融入路径与 D2C 商业策略在被问及 Dognosis 的犬类嗅觉检测技术如何融入传统医疗体系与保险报销网络时，Amir Alpert 首先为听众梳理了美国医疗保险运作的底层逻辑。他幽默地指出，现实中没有任何人会对“保险”这个词感到兴奋，更没有人愿意打电话给保险公司去争论自己的生命究竟值多少美元。然而，保险机构在美国及全球的医疗基础设施中扮演着核心角色。为了降低自身的审核与尽调成本，保险公司高度依赖权威监管机构与政策部门对药品和检测手段的官方背书与审批。针对这一漫长的准入流程，Amir 明确了 Dognosis 的阶段性战略：随着临床数据的持续积累，该检测项目未来终将纳入正规医保报销范畴；但在现阶段，团队已在合规与法律层面做足功课，率先开辟出一条直接面向消费者（D2C）的落地路径，以大众可负担的价格区间推向市场，并在推进过程中逐步完善后续的系统对接。 ```mermaid flowchart TD accTitle: Dognosis 市场准入与商业化路径 subgraph Phase1["阶段一：直接面向消费者 (D2C)"] A["完成法律与监管合规审核"] --> B["以亲民价格提供 D2C 检测服务"] B --> C["快速服务用户并积累真实世界临床数据"] end subgraph Phase2["阶段二：全面融入传统医疗保险体系"] C --> D["向监管机构提交充分验证数据"] D --> E["获得官方医疗审批与资质背书"] E --> F["正式纳入医疗保险报销体系"] end ``` - **保险公司的角色定位**：依赖外部监管机构的权威审批来决定覆盖范围，以规避逐个评估疗效的繁复尽调工作。 - **现阶段商业策略**：绕过漫长的早期医保审批周期，以直接面向消费者（D2C）的亲民定价提供检测服务。 --- ### [13:56](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=836s) 搜救与筛选犬只：从休斯敦救援犬 Snow 到印度的街头流浪犬主持人随后探讨了检测犬的选拔门槛，询问这是否像培育纯血赛马一样苛刻且难以匹配。Amir 借由现场的工作犬 Snow 揭示了检测犬极具包容性的来源。Snow 将成为 Dognosis 在美国首只专门训练用于癌症检测的工作犬，而它的前身是 Amir 七年前在休斯敦遭遇哈维飓风（Hurricane Harvey）后收养的搜救流浪犬。除了个别救援犬，Dognosis 目前在印度拥有一支由 15 只狗组成的检测犬队。这支队伍中虽然包含一部分历史上以嗅觉见长的小猎犬（Beagle），但相当大一部分同样直接来自印度的街头流浪犬。为了科学管理这群犬只，联合创始人 Ákos 甚至在周末利用 AI “氛围式编程”（Vibe Coding）快速开发了一款应用程序，通过社区协作的方式全面追踪这些犬只的健康与照料情况。Amir 强调，这项技术完全不需要“从 100 只狗中淘汰 99 只”，因为绝大多数普通犬只体内都潜在具备这种敏锐的嗅觉天赋，这极大地降低了项目的准入门槛与运营成本。 - **工作犬来源的多样性**：涵盖飓风救援犬（如现场的 Snow）、小猎犬（Beagle）以及印度的街头流浪犬。 - **AI 赋能犬只管理**：联合创始人利用 AI 工具在极短时间内开发了犬群健康追踪管理应用。 - **极低的筛选淘汰率**：无需像赛马那样极度挑剔血统，大多数犬类经过科学训练均能胜任气味识别工作。 --- ### [15:17](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=917s) 现场互动：Dognosis 自拍与大众意识传播在访谈进入尾声的最后五分钟，Amir 邀请全场观众参与 Dognosis 每场活动必不可少的传统环节——“全场合影自拍”。Amir 动情地提到，这不仅是本次活动的高潮，也是给首次搭乘飞机出差的 Snow 最好的回报与嘉奖。现场还发生了工作人员准备离场时被善意调侃的轻松插曲，引发一阵欢笑。合影之余，Amir 呼吁在场的听众拿出手机扫描大屏幕上的二维码，并在社交媒体上分享与标记 Dognosis。Amir 诚挚地向公众传递了团队的核心诉求：“我们不需要你的钱，我们需要的是你的认知与传播（We don't want your money. We want your awareness）。”他期盼大家在日常喝咖啡闲聊时，能够向身边的朋友讲述有一家创新公司正在利用犬类的力量检测癌症，并致力于为人类提供与犬类沟通的桥梁。只要大众愿意传播这份启发与认知，团队便将全心全意为大家服务。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=4s) 我是 Snow，Dognosis 的第一只狗狗。有任何问题尽管问我爸爸。虽然我现在还不会开口说话，但我听说公司正在紧锣密鼓地研发这项跨物种技术。 — Amir Alpert (以爱犬 Snow 视角) > [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) 就像得了一场流感一样 — Reed Jobs (Amir Alpert 引用) > [02:06](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=126s) 这才是真正的 AI 向善。一旦我们确立了这个使命，创造出的整体价值蛋糕将无可限量，到那时我们甚至会反思，当初为何要对金钱如此狭隘地执迷。 — Amir Alpert > [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) 无负担的多癌种早期筛查 — Amir Alpert > [05:00](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=300s) 具备嗅觉感知能力的 AI 硬件与底层架构 — Amir Alpert > [10:54](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=654s) 如同订阅流媒体服务一样无需犹豫 — Amir Alpert > [13:56](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=836s) 从 100 只狗中淘汰 99 只 — Amir Alpert > [15:17](https://www.youtube.com/watch?v=DRKW9zjtxRs&t=917s) 我们不需要你的钱，我们需要的是你的认知与传播（We don't want your money. We want your awareness） — Amir Alpert ## Mentioned Amir Alpert, Snow, Dognosis, Lurin, AI4, Reed Jobs, Yosemite, FDA, DogSense, Ákos, Hurricane Harvey ## Video description Amir Alpert - Head of US Business Development, Dognosis [Source](https://www.youtube.com/watch?v=DRKW9zjtxRs)

Source quotes

[0:04] Amir Alpert (以爱犬 Snow 视角): 我是 Snow，Dognosis 的第一只狗狗。有任何问题尽管问我爸爸。虽然我现在还不会开口说话，但我听说公司正在紧锣密鼓地研发这项跨物种技术。 [2:06] Reed Jobs (Amir Alpert 引用): 就像得了一场流感一样 [2:06] Amir Alpert: 这才是真正的 AI 向善。一旦我们确立了这个使命，创造出的整体价值蛋糕将无可限量，到那时我们甚至会反思，当初为何要对金钱如此狭隘地执迷。 [5:00] Amir Alpert: 无负担的多癌种早期筛查 [5:00] Amir Alpert: 具备嗅觉感知能力的 AI 硬件与底层架构 [10:54] Amir Alpert: 如同订阅流媒体服务一样无需犹豫 [13:56] Amir Alpert: 从 100 只狗中淘汰 99 只 [15:17] Amir Alpert: 我们不需要你的钱，我们需要的是你的认知与传播（We don't want your money. We want your awareness）

Source quotes

[0:04] Amir Alpert: I'm Snow. I'm the first dog in Dognosis. Any questions, just ask my dad. I'm not yet able to talk back, but I hear our company is working on that dog speed. [2:06] Amir Alpert: Focus on how AI improves the lives of the people that use it and of the lives that they can have when they're not using it... [2:06] Amir Alpert: Once we accept that the mission here is to improve the lives of everyone that interacts with AI, then we will see the value and the pie will be so big, we will sit in rooms like this and go, why did we ever question the money? [5:00] Amir Alpert: Multi-Cancer Early Detection & Triage [5:00] Amir Alpert: Smell Positioning System (SPS) [5:00] Amir Alpert: Olfactory AI [8:54] Amir Alpert: Will God be my witness? This dog has been with me through thick and thin. If I signed him up, I guarantee you we are treating these dogs with humanity that will inspire humans to be better humans to other humans. [8:54] Amir Alpert: DogSense [13:54] Amir Alpert: We don’t want your money. We want your awareness. We want you to tell your friends. We want you to get inspired. [13:54] Amir Alpert: Did you hear about the company that’s using dogs? Did you hear about the company that wants to give us the tools to talk to dogs? [13:54] Amir Alpert: You give us that and we are forever in your service.

Original transcript

[0:04] So this is the AI4 lens. Uh last session got cancelled unfortunately, but we're going to [0:08] crack on. This is going to be my favorite session of the whole event. So we get to choose something [0:13] that we really like that jumped out at us when we were programming the event. And this is what [0:18] this is what I chose for obvious reasons. Well, you know, not just because it's a cute dog, but [0:21] it's great technology that's hopefully going to change the world. So yeah, first question really [0:28] um you know we're here at AI4. There's a lot of enterprise AI there's lots of stuff who [0:34] knows what it's going to do but you know how you know tell me how this is a AI for good use case. [0:41] Oh before we get there we got to do introductions. Yeah we got to do introductions. Okay let's do it. [0:46] So Lurin I'm Lurin from AI4. Yeah. Amir Albert from Dnosis. Um, and star of the show, [0:55] Snow. You want to introduce yourself? I'm Snow. I'm the first dog in Dnosis. Any questions, just [1:02] ask my dad. I'm not yet able to talk back, but I hear our company is working on that dog speed. [1:08] Now we can start. [laughter] Very good. Great. So, thank you. Thank you for So, yeah, it it very much [1:17] is AI for good. But I know that's a bit of a loose phrase, but can you tell the guys about why why [1:23] your AI for good? How does that connect to a dog? It seems like a little bit of a disconnect. So [1:30] I think the thing is here, everyone at the conference is seeing like amazing technology [1:34] and everybody has that question. What is the future going to look like [1:38] when we take this technology and solve the world's biggest problems? Snow, the team at Dognosis, [1:46] what we're launching here at AI4, the official US launch of Dognosis HQ, is to set an example [1:53] to the smartest people in the rooms, the smartest people wandering these halls that were sitting in [1:58] here just a moment ago and say, "We have an understanding on how to apply AI for good." [2:06] That's what it is. Amazing. Amazing. And then before you get into the current technology as it [2:11] stands, I I think what's very important as well, can you tell everyone what your northstar is? [2:15] Again, an overused phrase. What's the long-term goal of your technology before we dive into the [2:21] nitty-gritty? I think today I'll share you guys some of the cancer work we're doing. Um, I think [2:25] it's a big statement, but I wouldn't say it if I didn't think it's true. I think Snow and the dogs [2:31] in our team are a big piece in solving cancer. And I don't use the word curing cancer. I'm not saying [2:37] people won't get cancer. I'm saying it won't hurt us as much as it used to hurt us. And I think to [2:44] borrow a quote from somebody who said it better than me, which is uh Reed Jobs, right? He started [2:49] Yoseite, this VC for cancer healthcare, and he says it should be like having the flu, [2:56] right? That's what we think snow and the team get us towards. Um and then the north star. Now, [3:03] you know, it's not a joke that I have Snow over here talking. The Northstar for dognosis [3:08] is interspecies communication. Apply that to dogs, your pets, the um horse rider in the race, [3:18] your cat. This is the beginning. Imagine a world where we no longer have to guess what's going on [3:26] because we can guess with education. We can guess with insight. we can guess with the guidance of AI [3:32] to help us understand what they're feeling, what they're thinking, and how to interact with [3:37] them in a more humane, but importantly partner relationship, not a pet. They escalate themselves [3:45] to be a partner and to wax a little on this. I think this is the overarching theme for AI. [3:51] Like the question everyone's asking, where's the value? Right? I'm sure you're thinking about that. [3:55] David Savage was thinking about that. Harvey Nash, right? Like what is the value of AI? [4:00] Right now the issue is that we're all thinking the value is money. How does the balance sheet [4:05] look bigger before AI after AI? It's a very real tech perspective. what we would suggest [4:14] to I hope this reaches the Elons and the Nvidias and you know I know we're a small company seed [4:20] round few million no big deal but to these people and I'll speak to you guys directly [4:27] focus on how AI improves the lives of the people that use it and of the lives that they can have [4:36] when they're not using it. Okay, we need to shift our perspective here. This isn't AI for money. [4:42] That's what AI for good is. Once we accept that the mission here is to improve the lives [4:47] of everyone that interacts with AI, then we will see the value and the pie will be so big, [4:54] we will sit in rooms like this and go, why did we ever question the money? Why were we so focused? [5:00] Sure. So just in terms of the technology, I know you're not the lead technologist at the company, [5:05] but can you explain in simple terms what the technology does from from dog to the [5:10] screen to Yeah, maybe a slide. Yeah, let's maybe this maybe slide will speak a thousand words. [5:25] Okay. So we think of dogs in terms of superpowers. Superpower number one that we're unlocking, [5:31] multicancer early detection that doesn't feel like a chore. What this means is you don't go [5:36] to a doctor's office you don't want to be at to make an appointment you didn't want to make cuz [5:40] your wife had to remind you of the appointment for 2 months if it's anything like my wife. [5:45] The kit comes to your door. You breathe into a breath mask, standard COVID N95 mask. [5:50] 10 minutes while reading the New York Times doing your Wordle puzzle. [5:54] And then you send that kid in, a trained pack of dogs guided by AI is able to determine [5:59] your chance of having cancer based on a K9 reactivity score. The reason they can do this [6:08] is because when cancer starts to form in the body, this is not a instant moment in time. This is a [6:16] progression of illnesses, of metabolic disorders, a lifelong journey from when you didn't have [6:23] cancer to when you did. The reality is that that lifelong journey is captured by the metabolict [6:29] activity in your body. The molecules that all of our tissues release into our bloodstream, [6:34] from our bloodstream to our lungs, from our lungs to our breath. And in our breath, canines [6:41] have a god-given tool to understand scent. What we're doing at Dognosis is fundamentally this. [6:52] We are closing the human human canine communication loop. We are removing the barriers [6:58] that stopped us from scaling by creating automated platforms that do the training and the sampling [7:04] and the testing and the recording. So it's no longer a subjective person making a decision. [7:10] It is trained quality controlled algorithms, quality controlled infrastructure. The same [7:15] standards that we hold for any diagnostic test, FDA approved drug will be applied to these dogs. [7:24] And maybe got to show you one of our dogs, Whiskey, doing its test. [7:31] Northstar though, because I not forget what the question was. [7:36] Superpower number one, cancer. In your brains, I want you to hear cancer and think healthcare. and [7:43] dogs for triage. What test do you need? The dog puts you into the right. Alzheimer's, Parkinson's, [7:49] blood disorders. The sky is unknown. If we can solve cancer, a disease that's been with us for [7:55] nearly 2 million years before humans even evolved to be called humans, as we call them today, [8:01] we can solve other diseases, too. Superpower number two is we're converting the smell of [8:07] the world into a visual map that a human can understand. We have GPS, [8:12] global positioning system. The north star for this superpower is a smell positioning system. [8:18] Where do we exist in the smell of s whatever word we want to come up with of the world? [8:25] And finally, maybe the most relevant to this audience when I said interspecies communication, [8:31] the fundamental hardware, software, and infrastructure layer of that is [8:34] AI that knows how to smell. At diagnosis, we are doing a reverse engineering project [8:40] that I won't get into too many details about, but the north star for us goes through this [8:45] AI that can do more than see. AI that can do more than speak. AI that can smell. [8:54] Finally, and I'll sit back down. We can get back to the fireside chat. That's all good. [8:58] This is Snow. He's going to be the first dog at Dognosis USA. He has been my dog for seven years. [9:07] I'm going to speak bluntly on this matter. When the topics of animals and science come together, [9:13] we have a history of maybe not making the right decisions as humanity. And when we post about [9:19] this on Instagram, there's always some people that say, "I hope those dogs are getting treated well." [9:25] Will God be my witness? This dog has been with me through thick and thin. [9:30] If I sign them up, I guarantee you we are treating these dogs with humanity that will [9:36] inspire humans to be better humans to other humans. So, please do not worry. [9:42] Finally, to end on a high note, this is Snow's brain. This is Snow's brain because we have this. [9:59] So to maybe get to one of your questions. All good. What is the stuff that we have [10:04] that will unlock this example number one is dog sense. This is a brain computer interface that [10:12] uses EEG technology that has been validated for decades on humans and we're applying it to dogs. [10:19] That means when Snow wears this, electrical stimuli from his brain come to the surface. [10:25] And when he's trained to be excited about something, like what we do with dogs in [10:30] the military for bomb mines or drugs at the border, apply the same principles to cancer. [10:36] We can see that excitement. In this example, I'm showing you his excitement to salami. Every one of [10:45] those spikes that you see on that screen, that's one moment in time that Snow got to smell salami. [10:54] Incredible. Incredible. So, it's great, I think, for us to understand the basics there. And as you [11:00] said, uh they can be trained for bombs and all these things we've we've known for for a long [11:05] time. Um so, it certainly makes sense in terms of, you know, the the cold hard business of it. [11:11] How does this scale? How do people interact with this? How do you, you know, how do you [11:16] embed in the infamous US, you know, insurance system, all that sort of stuff? Good question. [11:21] How do you normalize this? Good question. Um, who here has streaming services at home? [11:28] Okay. Uh, what's your name, sir? Sash, how much do you spend on your streaming services, [11:36] which is where we're headquartered in Bangalore, by the way. Yes. [11:39] All platforms put together. Yeah. All platforms put together [11:47] maybe $ 20, $30. So the average US household spends $77. Wow. [11:53] So I will tell you now that we don't have a price that's firm yet as we're still early in the US. [11:58] But this will be a price tag that if you don't think about your streaming services, [12:02] you won't think about the cost of getting a cancer check. Right? really fundamentally the thesis at [12:09] prognosis with regards to cancer. We all know what we should be doing, but damn it, life is hard. And [12:16] damn it, we don't always take care of ourselves. But in the age of AI with an abundance mindset, [12:23] what is there that we should be limiting ourselves with? For sure. One thing we talked about [12:30] earlier, sorry, I never answered your question. Sorry. Your question about health insurance. [12:35] How do you fit into that system? How would they validate this? you know in in Europe you can get [12:39] reflexology you can get all sorts of stuff but how would this fit into the traditional system here so [12:46] I have to pro provide a little context um and I provide this context out of the need to [12:53] maybe prevent what will come up in people's reactions when we talk about the insurance system [13:00] when I when somebody says the word insurance does anybody in here get excited does anybody want to [13:05] call an insurance company does anybody want debate why their life is worth some dollar amount. No, [13:12] but the insurance companies play a pivotal role in the US infrastructure and across the world. The [13:17] reality is because insurance has to do a good job, they require political bodies to approve drugs and [13:25] to give the vouch so that they don't have to spend all their time and diligence. Is this drug worthy? [13:30] That's a process. That process will take us time but eventually yes this will be covered once the [13:36] data accumulates. Until that time with clearance by legal and regulatory we did our homework. [13:43] We have a path to offer this direct to consumer at the priceless range that I mentioned [13:50] and we will work out the rest as it goes. Incredible. Uh one other question I had before [13:56] we we get the audience involved. We were talking earlier. I was curious about dog selection and you [14:02] know is it like uh purebred raceh horses will it be very difficult to match the best dogs and you [14:09] you had a great answer about how Snow came to this job. Do you want to explain to people the kind of [14:15] how many dogs this could be possible then? Yeah I mean Snow will be the first trained dog in the US [14:21] to detect cancer from diagnosis. I adopted Snow 7 years ago when I lived in Houston after Harvey. [14:29] So in India, we have a pack of 15 dogs. A lot of them are beagles. They have a [14:33] history of being good for this purpose. But a good number were from the streets. [14:40] You're from India, you know, street dogs in India, there's a lot. So we took them from [14:44] the streets and I shout out to our founder Akos. Akos Kolgard is the co-founder with Admir Britan. [14:50] Shout out to Akos because of the power of AI on the weekend. He vibe codes an app to help track [14:55] those dogs to give them the care they need in a communal way. Just vibe codes it over the weekend. [15:01] That's the power of AI. Brilliant. How does that show up in the balance sheet? Time will tell. [15:06] Sure. Yeah. Again, you don't need to get a 100 dogs, discard 99 of them. Yeah. You know, it's [15:12] it's a lower cost of entry to every dog or most dogs have this ability potentially within them, [15:17] which is fantastic. Yeah. So, we just have five minutes left. We thought we might uh invite people [15:23] to get involved. What What would you like to do? Yeah. So, open range. But before anybody [15:28] leaves the room, you walking out the room. Okay. You you work here. Good point. Uh um [15:37] at every one of our events, we had another talk on Monday. It was great. We want to do [15:42] the dognosis selfie. It's the cherry on top for almost coming out here and [15:47] snow traveling on his first plane ever to get here. The payoff for us is that [15:51] we get to take a selfie with the crowd and maybe one slide. Sure. [16:06] So, everybody, please, there's so there's 10 people in here. [16:11] Can we all just take out our phones and scan that QR code [16:21] and if you post about this, if you tell a friend about this, can you go on Instagram and tag us? [16:28] We don't want your money. We want your awareness. We want you to tell your friends. We want you to [16:33] get inspired. We want you to sit around coffee and say, "Did you hear about the company that's [16:37] using dogs? Did you hear about the company that wants to give us the tools to talk to dogs? [16:44] You give us that and we are forever in your service.