PaiX Navigator

PaiX Navigator is the flagship platform of the PaiX ecosystem, developed by PAICON. It enables users to query global health datasets in plain language, build patient cohorts across AI-ready public and proprietary data from 60+ countries, and either extract insights or license the underlying data — all within a single platform. It is designed to serve pharmaceutical companies, diagnostics developers, and research teams working across drug R&D, biomarker discovery, clinical trial design, and market analysis.

The platform sits at the top of a layered data infrastructure that includes PaiX Edge for ingestion and normalization at the point of care, and PaiX Atlas for longitudinal data harmonization. PaiX Navigator draws on this foundation to deliver harmonized, versioned, and audit-ready cohorts spanning multiple modalities and geographies.

Core Capabilities

• Agentic discovery: Users can ask questions in plain language; agents identify the relevant cohort and provide a transparent account of where every record originated.
• Multi-modal data support: Pathology and clinical data are available now, with radiology, omics, and lab data planned for future releases. All modalities are normalized to a single standard.
• Population-level insights: Biomarker prevalence and real-world outcomes are queryable across AI-ready public and proprietary datasets spanning 60+ countries.
• Disease coverage: Oncology is the primary focus, with 63 cancer types in the datalake. Cardiology, neurology, and rare diseases are in development.
• AI-ready datasets: Cohorts are labelled, versioned, and compliance-ready, designed to integrate directly into existing data pipelines.

Ways to Work with the Platform

• Data insights: Query the dataset behind a cohort and receive answers and briefs. The data itself never leaves the platform, making this suitable for market sizing and real-world evidence generation.
• Data licensing: License a cohort for in-house use under an annual license with a buyout option. Each licensed cohort ships with consent documentation and provenance records, making it audit-ready. Best suited for AI training and validation workflows.
• Data acquisition: If a required cohort does not exist in the datalake, PAICON will source it from hospital partners across its 60+ country network and deliver it as an AI-ready dataset. This pathway is particularly relevant for rare or underrepresented populations.
• Data partnership: Health data owners can become partners by hosting ethically approved, GDPR-compliant data on-site, retaining full control of the data while receiving a revenue share when their cohorts are licensed.

Licensable Cohort Examples

• PAI012 — Multi-cancer pathology cohort: Covers 2,606 patients with digitized whole-slide images (WSIs) paired with full clinical pathology reports. Includes 912 malignant patients with fully paired H&E and IHC slides and complete annotation. Total WSI count is 24,641 (8,950 H&E; 15,511 IHC). Annotations include diagnosis, tumour grade, specimen type, and IHC results. The deepest indication is breast cancer, with ER, PR, HER2, and Ki67 data across approximately 400 patients. Available under annual license with a buyout option.
• PAI034 — Global diagnostic consultation cohort: Approximately 72,000 real-world consultation cases submitted by clinicians worldwide, with approximately 927,000 QC-passed processed images. The slide mix is approximately 76% pathology (H&E, cytology, IHC) and approximately 2% radiology. Includes structured IHC data for approximately 4,700 cases across approximately 200 markers. Pan-organ coverage, with approximately 35% of classified cases malignant. Available under annual license with a buyout option.

Applications Across the Drug R&D Journey

• Biomarker discovery: Supports validation of biomarker prevalence and disease biology variation across geographies, rather than relying on cohorts skewed to a limited set of populations.
• Clinical trial design: Enables trial sizing, site selection, and stratification based on real cohorts from AI-ready datasets across 60+ countries, rather than literature or established site networks alone.
• Off-target and safety signal detection: Provides subgroup-stratified signals across regions and modalities before market launch, surfacing signals that may be too rare to detect in overall-population analyses.
• Patient stratification (Phase 3–4): Supports bias-stratified performance and outcomes analysis to quantify subgroup differences rather than reporting them as inconclusive.
• Market entry: Delivers real prevalence data by geography for market sizing and competitive intelligence, replacing modelled extrapolations from reference markets.

PaiX Navigator is positioned as a single environment for moving from an initial research question to a defensible, data-backed answer, with flexible options for how data is accessed, licensed, or contributed to the platform.