Aevius co-founders (from left to right): Long Phi Le, Rowland Pettit, Ming Yang (Max) Lu, and Faisal Mahmood. Photo credit: Aevius
Boston-based clinical intelligence company Aevius —a spinout from Mass General Brigham (MGB) —announced that it has raised $22M in funding led by global software investor Insight Partners to accelerate model development, team expansion, and enterprise commercial deployment.
Aevius was founded by Rowland Pettit, M.D., Ph.D. (CEO), Ming Yang (Max) Lu, Ph.D. (CTO), Faisal Mahmood, Ph.D., and Long Phi Le—key members of the founding team behind Modella AI, which was previously acquired by AstraZeneca.
Renowned physician-scientist and digital medicine authority Eric Topol, M.D., serves as Chair of Aevius’s Scientific Advisory Board, underscoring the platform’s commitment to clinical rigor and accuracy grounded in longitudinal patient histories.
Rather than relying on text-only large language models that reconstruct patient context one question at a time, Aevius trains multimodal AI models across structured EHR fields, unstructured clinical notes, and specialty medical data to generate a refreshable multimodal patient representation.
Rather than piecing together fragmented answers from unstructured progress notes one query at a time, Aevius develops foundational models trained across structured clinical tables, unstructured clinical text, and specialty multimodal medical data such as pathology, imaging, and genomics. By learning longitudinal relationships across a patient’s entire medical history, the platform constructs a refreshable multimodal patient representation—a living, computable mathematical state of the patient’s clinical condition and trajectory that updates dynamically as new encounters, diagnostics, and treatments occur.
Because this unified representation captures the whole patient rather than isolated data fragments, enterprise teams and downstream AI agents can build multiple clinical and operational applications on top of a single reusable foundation.
Source link







