Knit Health Raises $35M to Deploy Predictive AI
Knit Health raised $35M Series A led by NEA to deploy its Med-BEAM AI model for patient care prediction.

The big picture: Knit Health, a healthcare artificial intelligence startup, has raised $35 million in Series A financing. The round was led by New Enterprise Associates (NEA), with participation from Town Hall Ventures, SignalFire, and existing investors. This funding brings Knit Health’s total capital to $46.6 million and will support the commercial deployment of Med-BEAM, the company’s foundational AI model designed to predict patient care needs and evaluate potential clinical interventions.
Why it matters:
- Standardized Care: Knit Health aims to move healthcare beyond chance-based decisions by providing an ‘intelligence layer’ that predicts patient needs and the impact of interventions, making informed judgment available for every patient.
- Efficiency and Accuracy: Internal benchmarks indicate Med-BEAM significantly outperforms several frontier AI models in predicting clinical actions, demonstrating up to four times greater accuracy, 100 times faster processing, and 250 times lower cost.
- Broad Application: The Med-BEAM model is designed to power dozens of critical healthcare workflows, including care management, patient navigation, clinical trial recruitment, and hospital-at-home services, enhancing operational efficiency across various healthcare providers.
How it works:
- Healthcare ‘World Model’: Founded by University of California, Berkeley scientists, Knit Health is developing an AI system trained to understand patient trajectories by analyzing billions of clinical decisions and hundreds of millions of longitudinal medical records.
- Clinical Data Analysis: Unlike general-purpose large language models, Med-BEAM processes clinical actions—such as diagnoses, treatment orders, and referrals—to predict future care requirements and assess how different interventions might affect patient outcomes.
- Secure and Calibrated: The company ensures all training data is fully de-identified and uses only datasets with documented rights. The model also incorporates confidence calibration to identify predictions that may lack sufficient supporting evidence.
The catch: Knit Health’s claims of Med-BEAM’s superior performance are based on internal benchmarks, without independent validation or identification of the specific competing models used for comparison. This makes it challenging to objectively assess its competitive advantage in the crowded healthcare AI market, where numerous solutions are vying for commercial deployment across hospitals and health systems.
Key Facts
- Company: Knit Health
- Amount: $35M
- Round: Series A
- Investors: New Enterprise Associates (lead), Town Hall Ventures, SignalFire, Frist Cressey Ventures, Uncork Capital, Moxxie Ventures, Coalition Operators, The Aligned Fund
- Founder: Jonathan Kolstad
- Sector: Predictive Healthcare AI
AIPressRoom Weekly
Every round we reported this week, the people behind the companies, and the events coming up. One email, every Tuesday.