Knit Health's Behavioral AI Learns How Doctors Actually Think, Reshaping Patient Routes

A UC Berkeley spin-out just secured $11.6 million in seed funding to build something healthcare has never seen: an AI model that learns not from medical textbooks, but from how clinicians actually make decisions . Knit Health emerged from stealth in May 2026 with a Large Clinical Behavior Model (LCBM) trained on electronic medical record data from over 130 million patients across 30 U.S. health systems . While most healthcare AI predicts what might happen based on published literature, Knit's model observes what clinicians do—how they route patients, coordinate referrals, schedule follow-ups, and navigate institutional complexity—then applies those patterns to optimize care delivery . For healthcare marketers, this represents a fundamental shift: patient journey mapping may soon be driven by behavioral AI that knows where patients should go next before you do.

The seed round was co-led by Uncork Capital and Frist Cressey Ventures, with Moxxie Ventures leading the pre-seed and Coalition Operators participating . The company will use the capital to accelerate development and deployment of its behavioral model across health systems, focusing on clinical operations and patient care workflows . Knit trains its LCBM using deep reinforcement learning, causal inference, and behavioral cloning on Truveta EMR data—capturing the informal intelligence that determines whether a patient receives the right care at the right time . "Much of what matters most in medicine isn't written in textbooks, it's learned through experience with time and navigating the healthcare system," said Jonathan Kolstad, Co-Founder and CEO at Knit Health .

"Knit Health is creating a new approach to AI. Unlike traditional models, it learns and evolves from real human behavior and can be applied across complex systems," said Tripp Jones, General Partner at Uncork Capital . The distinction matters: traditional large language models generate probabilistic text based on patterns in published content, while Knit's behavioral model reflects observed sequences of clinical decisions—how care actually unfolds rather than how literature suggests it should .

This matters for healthcare marketers even if your organization isn't deploying behavioral AI today. If health systems begin routing patients based on AI-learned clinician patterns rather than traditional referral networks or patient preference, the tools marketers use to influence patient choice—SEO for condition pages, physician finder optimization, service line campaigns—may lose effectiveness. The patient journey becomes less about discovery and more about algorithmic assignment. Marketing leaders who assume patient acquisition strategies will continue to rely on search intent and content marketing may find themselves speaking to an audience that no longer controls the care pathway decision.

The Gap Between Clinical Intelligence and Patient Experience

Most healthcare AI applications focus on diagnostic support, administrative automation, or predictive analytics for readmissions and no-shows. Knit Health addresses a different problem: the collective intelligence embedded in how experienced clinicians navigate complex systems . This intelligence includes referral habits, capacity constraints, and the informal networks that determine which specialist sees which patient . These patterns drive outcomes because they represent accumulated institutional knowledge about who delivers the best results for specific conditions, which workflows reduce delays, and how to match patient complexity with provider expertise .

For marketers, this creates both risk and opportunity. Risk: if health systems deploy behavioral AI to optimize internal routing, patient preference may become secondary to algorithmic efficiency. A patient who researches orthopedic surgeons on your website may be routed to a different specialist based on capacity, historical outcomes, or scheduling patterns the AI learned from millions of prior decisions. Opportunity: marketers who understand how these systems work can align acquisition strategies with operational realities. If the AI routes complex cases to specific providers, marketing can build awareness around those providers' specialized capabilities. If the model optimizes for capacity utilization, patient acquisition campaigns can be timed to fill predicted gaps.

The company emphasizes that its model is health system-specific, fine-tuning to each organization's practice patterns, capacity constraints, and referral dynamics . This means the AI doesn't impose a generic best practice—it learns what works within a specific institutional context and integrates into existing operations . Navid Farzad, Managing Partner at Frist Cressey Ventures, framed the value proposition: "The hardest challenge in healthcare isn't knowing what good care looks like; it's delivering it consistently for every patient" . Knit embeds clinical intelligence directly into workflow to help clinicians make faster, more consistent decisions .

How Behavioral AI Changes Patient Acquisition Strategy

Traditional patient acquisition assumes patients exercise choice at multiple decision points: selecting a health system, choosing a specialist, deciding where to have a procedure. Behavioral AI collapses some of those decision points by automating routing based on learned patterns. A primary care physician's referral decision, historically influenced by personal relationships or patient requests, may instead be guided by an AI trained on outcomes data and scheduling patterns across the system . For marketers, this shifts the battleground from patient-facing content to clinician-facing education and internal stakeholder alignment.

Consider service line marketing for orthopedics. Historically, a marketer would optimize for patient search terms like "knee replacement surgeon near me," build physician profile pages with patient reviews, and run paid search campaigns targeting geographic areas. If a behavioral AI learns that patients with specific comorbidities achieve better outcomes when routed to surgeons with fellowship training in complex cases, the referral may bypass patient choice entirely. The patient researches surgeons online but the primary care physician—or the AI assisting that physician—makes the final routing decision based on clinical factors the patient doesn't see. Marketing must shift upstream: educating referring providers about specialized capabilities, ensuring the EHR data reflects accurate expertise, and aligning internal teams so the AI learns the patterns you want reinforced.

Knit's initial deployments target triage, patient flow, and quality improvement . These use cases directly impact patient experience and operational efficiency—both metrics healthcare marketers are increasingly measured against. Patient flow optimization affects wait times and appointment availability, which influence online reviews and net promoter scores. Triage decisions determine whether patients enter the system through emergency departments, urgent care, or scheduled appointments—each pathway with different cost structures and satisfaction profiles. Quality improvement initiatives shape outcomes data that drives physician rankings and service line reputation.

The funding Knit secured—$11.6 million for a seed round—signals investor confidence that behavioral AI represents a substantial market opportunity . For context, most healthcare AI companies at the seed stage raise between $3 million and $8 million. The size of Knit's round, combined with participation from Frist Cressey Ventures (a healthcare-focused investor with deep operational expertise), suggests the market sees behavioral models as differentiated enough to command premium valuations .

Follow the Money: What $11.6M in Seed Funding Reveals

Venture capital flows toward problems with large addressable markets and defensible technology. Knit's $11.6 million seed round indicates investors believe two things: clinical operations represent a massive efficiency opportunity, and behavioral AI creates a moat that traditional large language models can't replicate . The composition of the investor group reinforces this. Frist Cressey Ventures brings healthcare operational expertise and health system relationships . Uncork Capital provides early-stage venture experience and a track record with infrastructure companies . Moxxie Ventures and Coalition Operators add domain knowledge and network effects .

For healthcare marketing leaders, the investment thesis matters because it forecasts where health systems will allocate capital. If behavioral AI proves it can reduce length of stay, optimize bed utilization, or improve referral appropriateness, health systems will deploy it widely. That deployment changes how patients move through the system—and where marketing can influence those movements. A CMO who understands this shift can reallocate budget ahead of the curve, moving spend from patient-facing awareness campaigns to clinician education, EHR data optimization, and internal stakeholder alignment that ensures the AI learns favorable patterns.

Knit emphasizes HIPAA compliance, governance, bias testing, and continuous monitoring to ensure trustworthy guidance . These commitments address the regulatory and ethical concerns that have slowed AI adoption in healthcare. For marketers, this means behavioral AI is more likely to achieve broad deployment than earlier AI tools that raised privacy or fairness questions. The faster health systems adopt these models, the sooner traditional patient acquisition strategies face disruption.

The company positions its LCBM as "the infrastructure layer that every clinical decision runs on" . This ambition—becoming the invisible layer beneath routing decisions, discharge predictions, care team allocation, and referrals—would fundamentally alter how patients experience healthcare . If Knit succeeds, patient journeys will be shaped less by individual preference and more by system-level intelligence that optimizes for outcomes, capacity, and cost. Marketing's role shifts from influencing patient choice to aligning brand positioning with the care pathways the AI reinforces.

The 1ness Take

Healthcare marketers operate in a world where patient choice has been the organizing principle for two decades. Service line marketing, physician reputation management, and digital patient acquisition all assume patients research options and make decisions. Behavioral AI doesn't eliminate choice, but it does shift the locus of decision-making from patient to system. The question for marketing leaders: do you wait until health systems deploy these models, or do you start preparing now?

Our recommendation: start mapping how AI-driven routing would change your patient acquisition strategy. Identify which service lines depend on patient self-selection versus clinician referral. For high-referral service lines—complex surgery, specialized oncology, behavioral health—shift budget toward clinician education and internal data quality. If an AI learns referral patterns from EHR data, the accuracy of provider profiles, specialty designations, and outcomes reporting becomes a marketing asset. Work with IT and clinical leadership to ensure the data the AI trains on reflects your strategic priorities.

For service lines where patients exercise more direct choice—primary care, urgent care, elective procedures—behavioral AI poses less immediate disruption. But even here, triage optimization changes how patients enter the system. If an AI routes urgent care inquiries based on capacity and acuity, your call center scripts and digital scheduling tools need to integrate with that intelligence. Patient experience suffers when marketing promises immediate access but operational AI routes patients to the next available slot three days out.

The broader strategic shift: move from campaign-based marketing to system-based marketing. Instead of discrete initiatives—launch a joint replacement campaign, promote a new cardiologist, build awareness for a service line—think about how your brand shows up across the entire care pathway. Behavioral AI optimizes pathways, not touchpoints. Marketing must do the same. Map patient journeys end-to-end, identify where clinician decisions override patient preference, and build strategies that influence both. This requires deeper collaboration with clinical operations, revenue cycle, and IT than most marketing teams maintain today.

Finally, consider how behavioral AI changes competitive positioning. If two health systems in your market both deploy models trained on their own data, the system with better baseline patterns wins. The AI reinforces existing strengths and weaknesses—it learns from what clinicians already do. Marketing can't fix poor clinical performance, but it can ensure high performers are visible in the data. Physician recruitment, service line development, and quality improvement become marketing priorities because they shape what the AI learns.

The Takeaway

• Audit your patient journey assumptions. Identify which service lines depend on patient choice versus clinician routing. Behavioral AI will disrupt the latter first. Allocate budget accordingly.
• Invest in EHR data quality as a marketing asset. If AI learns from referral patterns, scheduling behavior, and outcomes data, ensure that data accurately reflects your strategic priorities. Work with IT and clinical leadership to audit provider profiles, specialty designations, and care pathway documentation.
• Build clinician education capabilities. As AI optimizes internal routing, marketing must shift upstream. Develop tools to educate referring providers about specialized capabilities, outcomes data, and capacity. The AI learns from what clinicians do—influence what they do.