Healthleap AI Raises $38M to Catch Hospital Risks Earlier

A hospital record can contain thousands of details about a patient, yet some of the most important warning signs can still be overlooked. Healthleap AI is betting that artificial intelligence can help hospitals find those signals earlier by reading both structured medical data and the clinical notes written throughout a patient’s stay. That strategy has now attracted $38 million in new backing as the company expands beyond its original focus on malnutrition.

The financing combines an $8 million seed round co-led by Sequoia Capital and First Round Capital with a $30 million Series A led by Hummingbird Ventures. Healthleap plans to use the capital across engineering, product development, sales, and customer success. The company is also preparing to extend its screening technology to a much wider range of hospital conditions.

Healthleap AI began with a considerably narrower mission. Siblings Jemima and Josiah Meyer founded the company after Jemima, a clinical dietitian, developed software to help identify patients who might be suffering from malnutrition. That early nutrition focus gradually exposed a much larger problem inside hospitals.

Important clinical clues often exist across different parts of an electronic health record. Laboratory results, vital signs, medications, weight, and diagnoses usually sit inside structured fields. Meanwhile, symptoms such as poor appetite, difficulty swallowing, muscle loss, confusion, or recent weight changes can appear inside written clinical notes.

Healthleap AI is designed to bring those signals together. Instead of waiting for clinicians to manually connect every detail, the platform analyzes inpatient records and assigns risk scores to people who may need additional attention. It does not diagnose those patients, but it can help medical teams decide who deserves a closer review.

That distinction is important. Healthcare AI attracts understandable scrutiny because mistakes can have serious consequences. Healthleap is positioning its system as a clinical prioritization tool rather than a replacement for medical judgment.

The company says its technology is now working with major healthcare organizations including Penn Medicine, Cedars-Sinai, Emory, Houston Methodist, Intermountain, Northeast Georgia, and UMass Memorial. Healthleap describes its platform as a way to screen adult inpatients every day and push risk information directly into existing hospital workflows.

Healthleap AI Is Turning Patient Records Into an Early Warning System

The basic problem Healthleap AI wants to solve is surprisingly familiar inside hospitals. Clinicians may have access to all the information required to recognize a problem, yet finding the right information at the right time remains difficult. A busy hospital can generate huge amounts of clinical documentation every day.

Healthleap attempts to reduce that burden by reviewing every adult inpatient record repeatedly. Its system analyzes information including laboratory results, vital signs, medications, weight changes, diagnoses, diet orders, and written clinical notes. It then writes a risk score into the hospital’s electronic health record so medical teams can see which patients may require more attention.

That workflow matters because clinical AI can become less useful when doctors and nurses need to open another application. Healthleap instead tries to place its recommendations inside systems clinicians already use. The company says its nutrition risk scores can appear directly in EHR flowsheets each morning.

Malnutrition became the company’s first major use case because it is both common and frequently underrecognized. Research has estimated that between 20% and 50% of hospital inpatients may be malnourished when admitted. Malnutrition has also been associated with longer hospital stays and poorer clinical outcomes in some patient groups.

Healthleap says the gap between risk and diagnosis can be substantial. Its website estimates that fewer than 9% of inpatients receive a malnutrition diagnosis, while a typical 1,000-bed hospital may miss thousands of affected patients annually. Those figures help explain why the company started with nutrition before moving into other conditions.

The startup has also put its malnutrition model through peer-reviewed validation. Healthleap says research covering more than 166,000 hospital admissions found its system had 88% higher sensitivity than the modified Malnutrition Screening Tool. Patients were identified an average of four days before their first dietitian documentation, according to the company’s published evidence summary.

Those results are significant because screening is only useful when it finds people early enough for clinicians to respond. A patient identified several days earlier could receive nutrition support before their condition deteriorates further. Earlier intervention may also affect the length and cost of a hospital stay.

Healthleap AI now wants to apply the same principle beyond nutrition. The company lists delirium and toxic metabolic encephalopathy, aspiration pneumonia, congestive heart failure readmissions, and pressure injuries among the conditions being developed or tested on the platform.

That expansion changes the company’s potential role inside hospitals. Instead of selling one tool to a nutrition department, Healthleap could become a broader clinical screening layer across the hospital. Its ultimate opportunity is to continually review patient records for risks that busy clinical teams may not see quickly enough.

Written notes are particularly important to that strategy. Many AI systems work well with structured data because numbers and standardized fields are easier to process. However, some of the richest clinical information exists inside the natural language doctors, nurses, dietitians, and other professionals use when documenting care.

A note might mention that a patient has barely eaten for several days. Another could record difficulty swallowing, recent weight loss, confusion, reduced mobility, or muscle deterioration. Individually, these details may appear ordinary, but together they can form an important risk pattern.

Language models give Healthleap AI a way to extract those concepts more systematically. The platform can look for whether a condition or symptom is actually present rather than simply finding a matching word. That matters because medical notes frequently include statements explaining that a symptom is absent.

The resulting information is then combined with structured clinical data. A patient’s laboratory values, weight, medications, vital signs, existing diagnoses, and written observations can all contribute to a risk assessment. The goal is not to produce an autonomous diagnosis but to prioritize the people clinicians should examine more closely.

That approach could become particularly valuable as hospitals face staffing pressures. Clinical teams cannot simply add unlimited employees to review growing volumes of information. AI systems that help prioritize attention could allow existing staff to spend more time with the patients most likely to benefit.

Healthleap says its deployments are already producing measurable operational results. At Cedars-Sinai, the company reports a 39% increase in malnutrition diagnoses without additional staffing. Healthleap also reports that affected patients experienced a 1.1-day reduction in length of stay.

Those figures are company-reported deployment outcomes rather than proof that the same results will occur at every hospital. Still, they explain why investors may see clinical prioritization as a commercially attractive use of AI. Hospitals care about patient outcomes, but they also operate under significant financial and staffing constraints.

Healthleap AI Is Building a Business Around What Hospitals Miss

Healthleap’s growth strategy is closely tied to proving that finding overlooked patients creates financial value. Hospitals are unlikely to buy another AI platform simply because its technology sounds impressive. Healthcare organizations increasingly want evidence that software improves care, saves staff time, or produces measurable economic returns.

Healthleap has structured its business model around that requirement. The company sells multi-year contracts based partly on the number of licensed hospital beds. It also ties elements of its commercial model to outcomes, giving hospital finance teams a way to compare the cost of the software with the value it creates.

One of its most prominent examples comes from the Hospital of the University of Pennsylvania. Healthleap reports $23.8 million in annualized financial impact from its program at that site. According to the company, $6.3 million came from additional reimbursement while another $17.5 million was linked to reduced length of stay and 8,632 annualized bed-days saved.

At Cedars-Sinai, Healthleap reports approximately $11 million in annual impact. The company also says every 1,000 additional reimbursable malnutrition cases captured can represent roughly $5 million in additional reimbursement.

Those economics reveal another side of the product. Healthleap AI is not only selling earlier clinical detection. It is attempting to demonstrate that better identification can improve hospital capacity and financial performance at the same time.

Shorter stays can create additional bed availability. More accurate documentation can also affect how hospitals are reimbursed for treating patients with greater clinical complexity. When those improvements are measurable, AI becomes easier for administrators to justify financially.

Healthleap says this approach helped it expand rapidly over the past year. Its careers site currently describes the company as a roughly 25-person team and reports 13-fold revenue growth. The company also lists more than $38 million in total funding.

The company’s hospital network has also grown from a handful of early partners into deployments across major health systems. Healthleap lists Cedars-Sinai, Emory, Houston Methodist, Intermountain, Northeast Georgia, Penn Medicine, and UMass Memorial among its partners.

For founders Jemima and Josiah Meyer, that expansion represents a major evolution from the company’s original product. Jemima began by developing a nutrition tool for fellow dietitians. Josiah brought the company-building side, eventually turning that clinical problem into a broader AI business.

Healthleap’s own description of its founding team captures that combination. It presents Jemima as the clinician who experienced the problem directly and Josiah as the builder who helped transform the idea into a company.

That founder-market connection may have helped Healthleap avoid one common healthcare AI problem. Technology companies can build technically impressive systems that fail to fit real clinical workflows. Starting with an issue observed directly by a dietitian gave Healthleap a specific problem to solve before it expanded its ambitions.

The challenge now is proving that its approach travels successfully from malnutrition into other medical conditions. Each new risk model brings different clinical signals, validation requirements, workflows, and consequences when the system gets something wrong.

Delirium, pressure injuries, aspiration pneumonia, and heart failure readmissions cannot simply be treated as different versions of malnutrition. Healthleap will need clinical validation for each use case and enough accuracy to prevent clinicians from being overwhelmed by false alerts.

Alert fatigue is already a serious concern in healthcare technology. Medical teams encounter numerous automated warnings inside electronic health records. If another AI system produces too many low-value flags, clinicians can begin ignoring the entire system.

Healthleap’s success may therefore depend on prioritization rather than maximum detection. Finding every possible risk sounds attractive, but hospitals need tools that identify the right patients with enough precision to justify clinical attention.

The company appears to recognize this. Its messaging emphasizes actionable prioritization rather than automated diagnosis. Healthleap wants clinicians to start their day knowing which patients deserve review first, not simply receive another long list of warnings.

The $38 million financing gives the startup more resources to build that broader platform. Engineering investment can help expand the models and improve integration, while customer-success teams become increasingly important as deployments spread across large hospital systems.

Sales will also matter because healthcare purchasing cycles are notoriously difficult. Hospitals evaluate security, clinical evidence, integration requirements, financial returns, and workflow changes before adopting new systems. Winning one health system can require considerably more work than selling typical enterprise software.

Healthleap has another advantage if its existing deployments continue producing useful evidence. Every successful implementation gives the company another case study when approaching other hospitals. Peer-reviewed research can also help separate clinically serious AI products from the flood of tools entering healthcare.

The ambition is ultimately much larger than malnutrition. Healthleap wants to identify risk across dozens of conditions and eventually extend its approach beyond inpatient hospitals into outpatient and home care.

If it succeeds, the company could become something closer to a continuous clinical safety net. Instead of waiting for a condition to become obvious, Healthleap AI would repeatedly examine the record and signal when scattered information suggests that a patient deserves another look.

That vision also captures one of the most practical opportunities for artificial intelligence in healthcare. The goal does not have to be replacing doctors or allowing software to make final medical decisions. Sometimes the more valuable role is simply making sure an important patient does not disappear inside an overwhelming amount of information.

Healthleap started with nutrition because that was the problem its founders understood firsthand. Four years later, the company is trying to prove that the same basic idea can extend much further.

The $38 million round gives Healthleap AI the capital to test that thesis at a larger scale. Its next challenge is showing that the technology can reliably find more of what hospitals are missing without creating another layer of noise for the people already responsible for patient care.