Referenced AI for healthcare
Creating trustworthy, steerable and impactful AI for healthcare
We work with organizations advancing science-based, equitable medicine. We’re building knowledge grounded AI to enable AI with control across practice and research.
Referenced AI x Generative AI = Trustworthy AI
Healthcare runs on evidence and explainability. Healthcare AI needs to too.
Large language models (LLMs), despite their strengths, are limited in support for the precision, explainability, bias awareness, timeliness, control and compliance healthcare needs. They need to be combined with grounded knowledge, planning and reasoning capabilities.
Referenced AI is a foundational AI system that integrates LLMs into an end-end architecture with these capabilities. Knowledge is ontologically-represented and atomically-cited to the peer-reviewed evidence, administration protocol or science that underpins it. All completely controllable by our customers.
Referenced AI x Generative AI is an architecture that leads toward more bias-aware, compliant and trustworthy healthcare AI.
Healthcare is at a mission-critical juncture
Clinicians and healthcare organizations are facing unprecedented pressures and mounting challenges. Demand is exceeding supply. Knowledge is fragmented and not effectively disseminated or accessible. There is information overload. The timeframes for new treatments to reach patients are protracted. And AI needs to be controllable.
That’s why we founded Evidium
To enable healthcare organizations across practice and research to harness scientific and clinical knowledge, and safely leverage the power of AI for the benefit of every patient.
Referenced AI Solutions
By combining Referenced AI and Generative AI, Evidium enables organizations to deploy medical knowledge in real-time and advance the connectivity between practice and research. We provide a data-centric platform where knowledge can be continuously assessed, applied and iterated on —safely, reliably and with referenced sources.
Building solutions for
clinicians.
clinicians.
-
Improve productivity, awareness, patient engagement, and quality
Enhance chart review, patient impact per visit and improve data quality at the front-end. Copilot is a Referenced AI assistant for use in clinical workflow. Chat in context with direct-line of sight to the knowledge. Access real-time insights and next-best-actions, integrated with EHR and ambient, with an "evidence finds me" experience.
-
Manage and scale AI-ready medical knowledge
Build, analyze, manage and deploy your own computational evidence and Referenced AI based solutions. Fusion is a Knowledge AI platform for scaling medical knowledge, administrative protocols and pre-clinical knowledge across practice and research use cases. Fusion comprises our clinical ontology, primed evidence graph, and customer-owned extensible graphs.
-
Enable semantic mapping of unstructured and structured data to knowledge
Semantically map your data to your own AI-ready knowledge. Enable multi-level, natural language exploration of clinical data combined with our computational medical intelligence language.
-
Kai-1 is a generalizable, knowledge-grounded AI interface for healthcare. Our API provides access to Referenced AI infrastructure for applications.
AI Engineering & Research
Referenced AI is an integrated system of neural and symbolic models, grounded directly on underpinning computational knowledge. Evidium undertakes applied AI research to solve healthcare specific challenges and opportunities. We’re building from first principles a unique end-end AI architecture with components including knowledge-grounded LLMs and a state-action model of medicine, to enable trustworthy and steerable characteristics. We have multiple issued and pending patents.
Come join us!
If you are interested in AI research engineering that will help advance health outcomes at scale, and drive toward safe AI in healthcare as a science-based environment, we’d love to hear from you.
Mission-Led
Our mission is to maximize the impact of medical knowledge and science, and make trustworthy AI available to every patient, practice and research organization. In so doing, we seek to support and amplify the work of medical experts and advance health outcomes for everyone.
We’re working toward a vision of creating Compliant Artificial Superintelligence for healthcare.
Evidium is a purpose-led AI company. Our strategic seed investors include Health2047, a Silicon Valley venture studio powered by the American Medical Association; and Interwoven VC, a specialist AI, robotics and healthcare investor.
Team
Carl Bate
Founder and CEO
Jennifer May Lee, MD
Chief Medical Officer
Thomas F. Unger, PhD
Chief Strategy Officer
David A. Epstein
Chief Product Officer
Matt McSorley
Lead Software Engineer
Hanna Roman, RHIA
Clinical Knowledge Engineer
John Harland, CPA
Chief Financial Officer
Advisors
Mark L. Graber, MD
Chief Medical Advisor
Founder and President Emeritus - Society to Improve Diagnosis in Medicine (SIDM)
Patricia Kingori, PhD
Ethics and Equity
Professor and Wellcome Trust Senior Investigator at University of Oxford
Victor Jakubiuk
Technology Advisor
Head of AI, Ampere
CTO & Co-Founder, OnSpecta
M Christine Stock, MD
Medical Advisor
Professor Emerita, Northwestern University Feinberg School of Medicine & Managing Director, Medical Affairs, Health2047, Inc
Michelle S. Keller, PhD
Research Advisor
Assistant Professor of General Internal Medicine, Health Services Research at Cedars-Sinai Medical Center
Board
Carl Bate, Founder and CEO, Evidium
Lawrence K. Cohen, PhD, CEO, Health2047
Lisa Chai, Managing Partner, Interwoven Ventures
Herb Madan, WGG Ventures
John Harland, CFO, Evidium
Schedule a demo.
Get on the Copilot waitlist.
Partner with Evidium.
Join our team.
Complete this form to get in touch.