Skip to main content
Minerva Clinical Intelligence
Minerva Clinical Intelligence
AI empowered clinical treatment
Minerva Logo

Where artificial intelligence becomes clinical wisdom.

Minerva Clinical Intelligence Ltd. is a clinical-stage AI company spun out of Imperial College London. Minerva's AI Clinician product line is composed of our unified Minerva Platform and Minerva AI Clinical Apps. The Minerva AI Platform is designed as regulated platform with pioneering clinical agents — collaborative AI  that recommends the treatment action, not just mere diagnostic or risk. Minerva AI Clinician Sepsis is the first app running on it build on the pioneering AI Clinician our team developed, trialed and deployed in multiple hospitals.


AI Clinician Platform

One EHR-independent runtime carrying the clinical data fabric, unified patient state, agent arbitration, safety and assurance, and the quality and regulatory system. Whether you are running EPIC, Cerner/Oracle or other EHRs we can operate on it.
Every app inherits it.

Read More


AI Clinician 
​Apps

Each app is a clinical agent for one specialty’s treatment problem. The clinically trialed Sepsis AI is first; we are currently developing subsequent medical use cases, that follow on the same layer, at falling marginal cost per app.

Read More


​Why Us



Spun out of Imperial College London by world-leading AI & clinician scientists,  we pioneered treatment AI

trained from operational data.

 With a decade of  peer-reviewed clinical evidence  (e.g. Nature Medicine, 2018, Nature Medicine 2019, npj Digital Health 2023, npj Digital Health 2024). 

Read More

Minerva Clinical Intelligence is a clinical-stage software company spun out of Imperial College London. We build the Minerva AI Clinician Platform & Agents: a regulated runtime on which clinical agents run. Healthcare AI has mostly shipped classifiers — software that predicts risk and stops before the treatment decision. Minerva ships agents: software that recommends the next treatment action, continuously, as physiology changes.


The hard part of that is not the model. It is the safety, assurance, regulatory and workflow layer a treatment-recommending device needs in order to be deployed at all. We built that layer once, as a platform, so that each new clinical app reuses it rather than rebuilding it. Our first app, Minerva AI Clinician Sepsis, recommends intravenous fluid and vasopressor dosing for adults in intensive care — the sharpest version of a problem that recurs across medicine.