Jemima Meyer was working as a clinical dietitian when she noticed the pattern. Patients at risk for malnutrition were arriving in the hospital already undernourished, getting missed on admission screening, and deteriorating before anyone looked closely at their charts. She built a tool to help. Her brother joined. They founded Healthleap in South Africa in 2022, and the company has now raised $38 million in seed and Series A financing from Sequoia Capital, First Round Capital, and Hummingbird Ventures to expand across U.S. hospitals.
The problem the company started with is stranger than it sounds. Up to half of hospitalized patients may be clinically at risk for malnutrition, according to Healthleap's figures, but fewer than nine percent receive a formal clinical diagnosis. The gap exists because the standard screening tool is a two-question survey administered once at admission, which misses sedated, intubated, and critically ill patients by design. What looks like a simple oversight turns out to be a structural failure of the intake workflow.
The surveillance model
Healthleap's solution is not a better screening questionnaire. It is a platform that sits on top of a hospital's electronic health record system and reads each patient's complete chart every morning, synthesizing physician progress notes, nursing assessments, laboratory panels, vital signs, medication orders, and active problem lists. It assigns risk scores and routes patients into existing clinical workflows for dietitians and care teams to act on. It does not diagnose. It flags.
A peer-reviewed study in Applied Clinical Informatics evaluated more than 166,000 hospital admissions over nearly four years. Healthleap's system showed eighty-eight percent higher sensitivity than the modified Malnutrition Screening Tool and identified at-risk patients an average of four days before the first dietitian documentation. At the Hospital of the University of Pennsylvania, the company reports $23.8 million in annualized financial impact, including $6.3 million in additional reimbursement and an estimated $17.5 million tied to 8,632 annualized bed-days saved.
The company has grown from three hospital partners to more than fifty in the past year. Its customers include Penn Medicine, Cedars-Sinai, Houston Methodist, Emory Healthcare, and Intermountain. Revenue increased more than tenfold in the same period, according to CEO Josiah Meyer.
The pattern it represents
Healthleap's arc follows a path that the broader clinical AI market has taken in radiology, ambient documentation, and now inpatient surveillance. Production wins have been workflow wins — not AI replacing the clinician, but AI ensuring the clinician sees the right patient at the right time. The FDA cleared close to 1,500 AI and machine learning-enabled medical devices through April 2026, roughly one per day in April alone. Most of them are in radiology. Healthleap's model suggests the next cluster will be in continuous inpatient monitoring.
The business case is straightforward enough that it passes the CFO's desk without translation: shorter hospital stays, higher reimbursement capture, measurable bed-day savings. That is a different conversation from the one that early clinical AI vendors had, when the argument was mostly that the technology worked. This one starts with what the technology costs and ends with what it pays back.
Malnutrition was the starting point. Healthleap is already expanding to other undetected conditions where the same daily-scan model applies. The architecture — ingest everything, score everyone, route to humans — is not specific to nutrition.
Topics healthaihospitalstechnology
