Primary source: Knit Health / Business Wire, November 16, 2026.

Knit Health Raises $35M for Med-BEAM. The Bet Is on Predicting Care, Not Just Automating Work.

Knit Health announced a $35 million Series A led by New Enterprise Associates on November 16, 2026, to deploy Med-BEAM, a model built to predict what happens next in a patient's care. This is not a new claims-processing tool. The RCM question is more interesting: what happens to the revenue cycle if software can anticipate the clinical decisions that create billing work in the first place?

$35 million

Series A led by NEA, with participation from Town Hall Ventures, SignalFire and existing investors. Knit reports $46.6 million raised to date.

Beyond large language models

Most healthcare AI applications focus on interpreting information, generating documentation or automating administrative workflows. Knit is targeting something different: modeling how healthcare itself operates.

Med-BEAM is trained on billions of healthcare decisions and hundreds of millions of longitudinal patient records, according to the company. It learns patterns in diagnoses, referrals, orders and treatment decisions to predict likely care trajectories and compare how interventions might change them.

Knit says its technology can serve health systems, life sciences companies, healthcare technology providers and other organizations. The company was founded in 2025 by scientists from the University of California, Berkeley.

Read the benchmark carefully

In a company-run test predicting clinical actions that hospitalized patients would take during the next 24 hours, Med-BEAM was up to four times more accurate, 100 times faster and 250 times less expensive than the frontier language models compared. Knit also reports better calibration between prediction confidence and observed outcomes.

These are vendor-reported results on a specific clinical prediction task. They are not independent evidence that Med-BEAM improves clinical outcomes, reduces claim denials or boosts collections. The announcement did not disclose RCM deployment results.

Why revenue cycle leaders should care

RCM teams inherit upstream decisions. A referral creates an eligibility and coverage question. A treatment plan creates authorization and medical-necessity requirements. A service performed in one setting instead of another changes coding and reimbursement. Billing teams then spend hours reconstructing why those decisions were made.

RevCycleAI's read: A sufficiently reliable predictive model could eventually push revenue-cycle work earlier, from reacting to claim errors to anticipating the operational choices that produce them.

  • Prior authorization: Anticipating likely services could give patient access teams more time to secure approvals.
  • Documentation and coding: Expected care pathways could surface documentation requirements before the chart reaches coding.
  • Denial prevention: Combining clinical prediction with payer rules could flag reimbursement risks before claim submission.
  • Revenue forecasting: Likely care utilization could inform capacity and revenue estimates earlier.

Those are potential applications of predictive clinical intelligence, not features Knit has announced for its Med-BEAM product. Predicting the next clinical action is not the same as predicting payer reimbursement.

The bigger shift: from automation to intelligence

Healthcare AI's first investment wave has concentrated on doing existing work faster: writing the note, coding the chart, contacting the payer, drafting the appeal and collecting the balance. Knit represents a different bet. It seeks to understand the system generating those tasks.

If one specialized model becomes reliable across multiple healthcare decisions, many applications could eventually run on top of that intelligence. The competitive question changes from who automates an individual workflow to who owns the data and predictive capabilities that power multiple workflows.

The RCAI signal

$35 million is funding a model of future clinical decisions, not a replacement for billing software. The longer-term RCM opportunity lies upstream: anticipating the authorizations, documentation requirements and payment risks created by a care decision before they hit the back office.

RCAI take

Knit Health's financing reinforces a developing healthcare AI investment thesis: the potential value is shifting beyond labor automation toward specialized intelligence. Most RCM systems respond after an event happens. Predictive clinical models could eventually help teams anticipate the events that create administrative work.

The next evidence to watch is a demonstrated link between those predictions and measurable authorization, denial or financial results. That link is not in the announcement. But it is where this technology could become meaningful to revenue-cycle operators.

Source: Knit Health via Business Wire, November 16, 2026. Fundraising, model scale and benchmark figures are company-reported. RCM applications are RevCycleAI analysis, not announced Med-BEAM product functionality.