Leo Anthony Celi is the principal investigator behind the Medical Information Mart for Intensive Care (MIMIC) and its derivatives (MIMIC-CXR, MIMIC-ED, MIMIC-ECHO, and MIMIC-ECG), serving over 100,000 users worldwide and generating over 10,000 publications that have fundamentally shaped machine learning in healthcare. His group at MIT leads MIT Critical Data, a global consortium whose mission is to build local AI capacity and agency in healthcare through open data and software, foster community engagement through datathons and workshops, and advance equity through technology. His work pioneers process-based approaches to algorithmic fairness that move beyond traditional demographic metrics, developing culturally-aware human-AI systems that incorporate diverse knowledge traditions including indigenous epistemologies and faith perspectives. In partnership with hospitals, universities, and professional societies worldwide, Dr. Celi's team has organized over 100 health data science events across more than 40 countries, democratizing access to opportunities for growth and fostering interdisciplinary collaboration that actively addresses power dynamics in the global knowledge ecosystem.

More recently, Dr. Celi's team has developed DOJO (Distributed Open Justice Oversight), a community-driven adversarial evaluation platform for health AI, alongside complementary frameworks including BODHI, an epistemic virtue framework for clinical AI; SAKSHI, an investigative accountability platform modeled on journalism; and LTARC, a scaffold for local, task-specific, community-partnered evaluation. As Editor-in-Chief of PLOS Digital Health, he champions structural reform in scientific publishing. Their "AI as a Catalyst" workshop series brings together indigenous knowledge holders, artists, activists, and patients alongside clinicians to reshape medical education and workforce development.