Paige is a primary example of specialized vertical AI, building what can be characterized as 'expert agents' for the medical domain. While most current agent discourse focuses on LLMs navigating web interfaces or general business tools, Paige is developing autonomous systems that navigate complex biological data. Their work on foundation models for pathology provides the cognitive architecture for future agents that could screen slides, draft diagnostic reports, and suggest treatment paths based on tissue morphology.

Within the agent ecosystem, Paige represents the transition from narrow computer vision to agentic systems that understand biological context. Their collaboration with Microsoft to build the world's largest image-based AI model for cancer is a fundamental step in creating clinical agents that can function with a high degree of autonomy in diagnostic workflows, pushing the boundaries of what specialized AI can achieve in high-stakes environments.

The digitization of the biopsy

Paige is attempting to do for pathology what radiology underwent decades ago: a transition from analog glass slides to a fully digital, software-driven workflow. For over a century, the pathologist's work has been tied to the physical microscope. Paige enters this market with the premise that AI can digitize these samples and provide a layer of analysis that exceeds human throughput and consistency. The company’s core offering is a clinical-grade platform that allows labs to manage digital slides, paired with intelligence modules that scan these images for signs of malignancy.

The system operates by converting physical tissue samples into high-resolution digital images. Once ingested into the Paige Platform, algorithms—such as the company's prostate and breast cancer detection tools—examine the tissue at a cellular level. These are not intended to replace the clinician but to act as a co-pilot, flagging suspicious areas for closer inspection. This human-in-the-loop approach is central to how Paige navigates the clinical environment, ensuring that the AI remains a supportive tool for expert decision-making.

Foundation models for tissue morphology

While many healthcare AI companies focus on narrow, task-specific algorithms, Paige is pivoting toward a foundation model strategy. In partnership with Microsoft, the company is training large-scale models on petabytes of de-identified image data. This effort, often associated with their Virchow model, aims to move beyond simple cancer detection. By training on millions of slides across multiple organ systems, the company is building a general-purpose understanding of human tissue morphology.

This shift mirrors the evolution seen in natural language processing. Instead of building a new model for every specific cancer type or tissue variation, a broad foundation model can be adapted to various clinical tasks with minimal additional training. This capability is critical for scaling across the vast variety of diseases that a general pathology lab encounters, potentially making the system useful for everything from rare oncology cases to routine diagnostic screenings.

Regulatory moats and clinical reality

Unlike general-purpose AI startups, Paige operates in a heavily regulated environment where the stakes are literal life and death. The company achieved a significant milestone by receiving the first-ever FDA de novo marketing authorization for an AI-based software to identify cancer in prostate biopsies. This regulatory approval is more than a legal requirement; it is a competitive moat that separates clinical-grade tools from research projects.

The competitive landscape is divided between legacy hardware manufacturers like Philips and Leica, who are adding software to their hardware stacks, and pure-play AI competitors. Paige's advantage lies in its deep integration with world-class clinical data from Memorial Sloan Kettering and its early lead in the regulatory process. As hospitals move toward operational efficiency, Paige’s business model—offering the viewing platform as a SaaS layer with per-module diagnostic fees—aligns with the digital transformation currently sweeping through medical systems globally.