“Pangaea’s discovery of more undiagnosed, misdiagnosed and miscoded patients based on clinically actionable intelligence is helping healthcare providers improve patient outcomes, manage resources and collaborate with pharmaceutical companies. As a result, their customers can advance research across rare and hard-to-diagnose conditions without the need to transact data and thereby comply with privacy regulations.” – Kevin O’Leary, Regional Business Leader, Health & Life Sciences, Microsoft

Pangaea Data’s AI platform, PALLUX, is configured on clinical guidelines to help healthcare organizations (HCOs) and pharmaceutical companies find more untreated (undiagnosed, miscoded, at-risk) and under-treated patients, who are currently missed despite information available in their records.

Take 55-year-old Mr. Smith, for example. He visited his clinician for shortness of breath and a persistent cough and received a diagnosis of asthma with an inhaler prescription. Despite treatment, his condition continued to worsen. Why is this? In reality, clinicians often have only 10 minutes per consultation. Even with over one-third of their time spent on manual record review, patient records can span hundreds of pages, making it nearly impossible to catch every critical detail. In Mr. Smith’s case, a closer inspection of his records reveal chronic cough and sputum production lasting over three months each year for the past three years — indicators for Chronic Obstructive Pulmonary Disease (COPD) screening per the GOLD guidelines.

PALLUX addresses these gaps by analyzing patient records, including both structured and unstructured notes. The platform goes beyond conventional NLP approaches by considering context and applying reasoning based on clinical guidelines, effectively mimicking a clinician’s manual review. This enables PALLUX to accurately flag patients with hard-to-diagnose conditions in real-time at the point of care, ensuring that patients like Mr. Smith receive timely, guideline-concordant screening and treatment, reducing delayed or missed diagnoses.

PALLUX is configured for new conditions within 12 weeks and is deployed behind the HCO’s firewall, ensuring privacy compliance. The platform seamlessly integrates into existing electronic health record (EHR) and scheduling systems, as well as with ambient listening tools, ensuring no clinical or IT workflows are disrupted. Furthermore, as a platform, PALLUX allows HCOs to integrate their existing AI models, enabling their application from both a point of care and population health perspective, while providing clinicians with a single dashboard to access all insights.

For example, a global pharmaceutical company has deployed PALLUX to help find 5 times more COPD patients and 2 times more COPD patients at high-risk of exacerbation. This is helping to close care gaps and ensure that these thousands of patients who may otherwise have been missed receive timely diagnoses and appropriate treatments, ultimately expanding the addressable patient population. This has a ripple effect on revenue streams, reducing later costly, complex care, increasing uptake of medications and advanced therapies while also improving recruitment for clinical trials, which often struggle to enroll patients in early disease stages.

PALLUX has demonstrated results across disease areas, including but not limited to: pulmonology (e.g. COPD), nephrology (e.g. CKD), oncology (e.g. Lung Cancer, Ovarian Cancer, Breast Cancer, Cancer Cachexia), autoimmune (e.g. SLE, Lupus Nephritis), and more (e.g. Primary Biliary Cholangitis, Herpes Simplex Virus).

Pangaea Data is a South San Francisco and London based company. Its founders, Dr. Vibhor Gupta and Prof. Yike Guo, have raised over $300 Million in research funding, including a $110 Million grant for the development of generative AI in medicine through which they have procured 1,500 GPUs from NVIDIA. Their investors and advisors include industry veterans, such as Lord David Prior (Former Chairman of NHS England) and Mr. Andy Palmer (Former CIO, Novartis).

To learn more, please contact us at info@pangaeadata.ai