Slideflow Labs

Market position

The company operates in AI-driven digital pathology and computational biomarker discovery, positioning itself against centralized, high-cost legacy molecular tests by running validated models inside a hospital's existing pathology and scanning infrastructure. Leadership frames the objective as democratizing access so that advanced AI pathology is not limited to large companies or institutions with extensive machine-learning expertise, and states a goal of exceeding the performance of widely used legacy tests.

Local, single-tenant deployment inside the customer institution's own environment, so health systems retain control of patient data while running advanced AI models; use of the standard H&E slide with no additional stains or specimen shipping; near-real-time turnaround delivered into existing clinical reporting workflows; support for low-cost and varied scanning hardware; the ability to accept models built by other teams to enable collaboration between medical centers; and an open-source, peer-reviewed foundation cited or extended by more than 100 studies.

Technology

Slideflow Labs' platform builds on Slideflow, an open-source, end-to-end deep learning library for digital histopathology originally developed by James Dolezal and collaborators. The software supports whole-slide image processing and tiling, stain normalization, weakly and strongly supervised model training, multiple-instance learning, self-supervised learning, foundation models, explainability tools, uncertainty quantification, and an interactive deployment interface; Slideflow Studio provides real-time whole-slide predictions and heatmaps. Underlying research includes computer vision models that predict genomic alterations, gene expression signatures and molecular subtypes directly from routine histology slides, as well as methods to improve trust in model outputs and to correct site-specific confounding patterns that reduce accuracy. The team is also developing quality-control methods intended to make AI outputs reliable and appropriate for patient-facing use.

Go-to-market

Direct engagement with hospital clinical and pathology teams, with deployment fitted into existing scanning infrastructure and IT/security review processes. The company has pursued pilot programs with biotech, pharma and healthcare partners and has used university and accelerator channels — the Polsky Center Transform accelerator, Alchemist Chicago, and the Northwestern Medicine & Techstars Healthcare Accelerator — for clinical access, commercialization support and investor networks. The open-source framework and a large body of peer-reviewed publications serve as a funnel and credibility channel for researchers and clinicians.

Hospitals, health systems and academic medical centers with pathology and digital slide-scanning operations, including institutions lacking large-scale digital pathology infrastructure or in-house machine-learning expertise; also biotech and pharma partners for pilot programs, and academic and clinical research teams using the open-source framework.

Geography

Headquartered in Chicago, with deployment at academic medical centers including the University of Chicago; the platform is described as deployable to hospitals across the country, and the open-source framework is used in research worldwide.

History

Slideflow Labs commercializes intellectual development from the University of Chicago. In February 2025 it was profiled as a participant in Cohort 4 of Transform, the Polsky Center and Data Science Institute data science and AI accelerator, with near-term plans to complete its web interface and cloud integration, launch pilots with biotech, pharma and healthcare partners, and begin the FDA regulatory compliance process. In October 2025 it was listed in the inaugural Alchemist Chicago cohort and in January 2026 advanced to Alchemist Chicago Phase 2/Class 41. In March 2026 it joined the 2026 Northwestern Medicine & Techstars Healthcare Accelerator, and on June 5, 2026 it won first place and a $575,000 prize at the 30th annual New Venture Challenge. Stated next milestones include expanding deployment, validating performance across institutions and advancing regulatory pathways for clinical use.

Risks & controversies

The company's clinical products are not described as having completed regulatory clearance; it stated in 2025 that it was beginning the FDA regulatory compliance process, and 2026 reporting lists advancing regulatory pathways for clinical use as a future milestone. Published team research also addresses site-specific digital histology signatures that can reduce deep learning model accuracy and introduce bias, an acknowledged technical risk the company says it works to correct.