Knit Health has emerged from stealth with $11.6 million in seed funding to advance an artificial intelligence platform designed to improve healthcare operations and clinical decision-making. The San Francisco-based company, spun out of the University of California, Berkeley, is developing what it calls a Large Clinical Behavior Model, or LCBM. The seed round was co-led by Uncork Capital and Frist Cressey Ventures, following a pre-seed round led by Moxxie Ventures with participation from Coalition Operators.
Building AI Around Real Clinical Behavior
The company is positioning its technology as a response to a major limitation in healthcare AI: the reliance on language models trained mainly on medical text, published research, or documentation. Knit Health argues that much of what determines effective care happens outside formal records, including referral patterns, scheduling decisions, care coordination habits, and the practical judgment clinicians develop inside complex health systems. By learning from these real-world behaviors, the company aims to help providers route patients more effectively and support faster, more consistent care delivery.
A Model Trained on Large-Scale Clinical Data
Knit Health’s LCBM is being trained on Truveta EMR Data covering more than 130 million patients across 30 U.S. health systems. The company says the model uses deep reinforcement learning, causal inference, and behavioral cloning to identify patterns in how clinicians make decisions over time. Rather than generating text-based recommendations alone, the platform is intended to model how care actually unfolds across patient journeys and operational workflows.
Focus on Health System Operations
The funding will support further development and deployment of Knit Health’s technology across health systems, with early use cases including triage, patient flow, discharge prediction, quality improvement, and care team allocation. The company says its system can be fine-tuned to individual health systems, reflecting local practice patterns, capacity constraints, and referral dynamics. This health system-specific approach is intended to make the technology more practical for real clinical environments where operational realities often shape patient outcomes.
Investor Backing and Market Rationale
Jonathan Kolstad, Knit Health’s co-founder and CEO, said the company is focused on capturing the knowledge clinicians gain through experience navigating the healthcare system. Investors backing the company framed the approach as a new category of AI infrastructure that learns from human behavior rather than relying only on static clinical content. Uncork Capital’s Tripp Jones and Frist Cressey Ventures’ Navid Farzad emphasized that Knit Health could help scale clinical intelligence across complex organizations and make high-quality care more consistently available.
Governance, Compliance, and Deployment
Knit Health says its platform is being built with HIPAA compliance, governance controls, bias testing, and continuous monitoring. These safeguards are central to the company’s effort to build trust with both providers and patients as its models are introduced into healthcare workflows. The company is currently partnering with health systems to deploy initial models and validate their role in improving care coordination and operational performance.
Founded in 2025 by UC Berkeley researchers and academics, Knit Health is entering the market with a clear thesis: healthcare AI should learn not only from medical knowledge, but also from the collective behavior of clinicians. Its $11.6 million seed financing gives the company capital to expand development and pursue real-world deployments across health systems. As hospitals and providers continue searching for ways to improve efficiency, consistency, and patient outcomes, Knit Health’s clinical behavior model could become a notable addition to the next generation of healthcare AI infrastructure.