The digital innovation team at Zuckerberg San Francisco General Hospital.

Pioneering Research and Organizational Solutions to Promote Equitable Care through Technology

We apply technology and digital tools to improve health outcomes and equity in vulnerable and underserved populations, and we deploy what we build inside the hospital's own electronic health record.

Read the ZSFG Predictive Analytics Report

Systems in daily use across the hospital, built with clinicians and operational leaders in the Kaizen Promotion Office.

27.9% to 23.9% 30-day heart failure readmission rate

Heart failure

Standardized inpatient heart failure care with AI and EHR automation. The initiative reduced readmission rates, reduced heart failure mortality rates and closed equity gaps in care between Black/African American patients with heart failure and the general heart failure population.

100+ a year hospital admissions avoided

Chest pain pathway

A decision-support pathway in the emergency department, with patients evaluated faster and with fewer downstream tests.

4 times more likely to receive timely specialist care

Valve disease

AI identifies severe valve disease and routes patients to treatment.

89% agreement with human reviewers of the same charts

Whole Person Navigator

AI that summarizes a patient's medical and social history for social workers. The 23 inpatient social workers at ZSFG together spend more than 300 hours a week reviewing charts before they see patients, and the Navigator gives that time back to patient care. In validation it matched human reviewers as often as the reviewers matched each other.

See all projects

Clinicians, engineers and data scientists working inside the hospital.

Meet the team

Lucas Zier Jean Feng James Marks Hemal Kanzaria Christopher Peabody Robert Gallo Patrick Vossler Jialin Ouyang

Institutional home

Zuckerberg San Francisco General Hospital and Trauma Center University of California San Francisco San Francisco Department of Public Health ZSFG Kaizen Promotion Office

Collaborators

ShelterTech

Recent news

All news

January 2026

Our team is looking for a (senior) AI engineer. If you are interested, please check out the posting.

December 2025

Our research on how AI could help social workers review patient records was presented at the NeurIPS GenAI4Health workshop and accepted for oral presentation at AAAI. Read the paper.

December 2025

Work on building compact, transparent AI models for healthcare was accepted at NeurIPS. BC-LLM learns fully interpretable risk prediction models clinicians can use at the bedside. Read more.