Healthleap Secures $38M to Expand AI‑Driven Patient Risk Detection

Technology and Approach

Healthleap’s solution pulls data from both structured fields and free‑text clinician notes inside a hospital’s electronic health record system.

Using language models, it extracts affirmative or negated mentions of concepts such as recent weight loss, poor appetite or difficulty swallowing.

These signals are combined with lab results, vital signs and medication data to produce a risk score that is written back into the care team’s workflow each morning.

More than fifty hospitals now run the platform, including Penn Medicine, Cedars‑Sinai and Houston Methodist.

Healthleap reports a ten‑fold increase in revenue over the past year and offers three‑year contracts priced by licensed bed count, with an outcome‑based option.

Implications for GPU‑Powered AI Infrastructure

Running natural‑language inference on thousands of patient charts each night demands scalable GPU acceleration.

Healthleap’s workload highlights the growing need for high‑throughput AI servers that can handle mixed structured and unstructured data in real time.

For providers like AiGpu, this translates into demand for GPU instances optimized for low‑latency language model inference and secure, HIPAA‑compliant environments.