HealthLeap

HealthLeap builds an AI-powered malnutrition screening safety net for hospitalized patients. The company's tool draws on hospital electronic health record (EHR) data - labs, vitals, clinical notes and structured records - to generate real-time risk scores and flag conditions that would otherwise go unnoticed. It screens 100% of hospitalized patients daily and processes billions of data points from hospital EHR systems.

The clinical problem is substantial: between 20% and 50% of hospitalized patients are malnourished, yet fewer than 9% of malnourished patients are diagnosed at baseline. HealthLeap describes its AI Malnutrition Screening Tool as the only validated malnutrition AI screening tool on the market. During patient stays the model achieves an AUROC of 95%, delivers 88% higher sensitivity than manual nurse-administered malnutrition screening, and identifies malnourished patients four days earlier. Institutions using the platform include Cedars-Sinai, Penn Medicine, Houston Methodist, Emory and Intermountain Health. The company operates in the United States.

Work at HealthLeap spans artificial intelligence, machine learning, EHR data processing, clinical risk scoring, natural language processing of clinical notes and predictive analytics. The team is registered-dietitian-led and was founded by siblings; the company is backed by Sequoia and First Round Capital. Its stated ambition is to extend the approach from malnutrition to every major condition.