Ichilov Hospital completed an unprecedented move on August 12: migrating the entire infrastructure of its Chameleon medical record system to the AWS cloud, including about 170 internal and external interfaces. The move marks the first step in a three-year plan to transform Ichilov into an AI-first hospital, one capable of rapidly deploying advanced tools, from automated visit documentation to generating surgery summaries at the touch of a button.
The transition to cloud infrastructure reflects a global trend reshaping entire sectors. Banks, insurers, transportation companies, and e-commerce platforms already rely heavily on the cloud to process massive data volumes and maintain operational flexibility. In healthcare, the shift has been slower due to data sensitivity and strict regulation. Yet the coronavirus pandemic accelerated adoption: leading institutions such as the Mayo Clinic in the U.S. and the NHS in the U.K. have begun moving core systems to the cloud to integrate artificial intelligence into diagnosis and treatment.
AWS, Amazon’s cloud computing division, allows organizations to rent computing and storage capacity on demand. In August 2023, AWS established three local server farms in Israel under the government’s Nimbus project, serving as the country’s cloud infrastructure backbone.
“We Will Free Doctors From Administrative Burdens”
Most hospitals in Israel have until now deployed limited digital solutions such as online appointment scheduling or partial data analysis. Ichilov’s migration of its core electronic medical record system, with its dozens of interfaces, is precedent-setting in Israel and among the first of its kind worldwide.
For Ichilov, the rationale is twofold: local infrastructures are insufficient for the exploding volumes of data and AI workloads, while cyberattacks and even rocket fire near hospitals have underscored the need for resilient, distributed systems.
“The daily lives of medical teams are complex and Sisyphean, so making data accessible and processing information for decision-making is critical,” says Yariv Nir, Deputy Director, Technologies and IT, at Ichilov. “The move to the cloud is designed to simplify processes, free doctors from administrative burdens, and restore attention to the patient encounter.”
He describes an AI tool under development: “To release a premature baby after 100 days in the hospital, it currently takes a full day to write a summary. With AI, this can be done in three minutes, leaving doctors with more time for patient care.”
Another example is speech-to-text: “Doctors can speak, the system summarizes automatically, and with a click, a complete report is produced. This way, doctors look at patients instead of screens.”
Toward More Personalized, Efficient Care
Critics often warn of over-reliance on a single cloud provider, raising concerns about cost, flexibility, and service continuity. Nir argues that cloud-based distribution enhances resilience: “When missiles landed 200 meters from Ichilov, the risk of having only a local data center became clear. In the cloud, we have full backup and a multi-cloud strategy in development.”
The adoption of AI in medicine also raises challenges of liability, bias, and training. Ichilov has established a dedicated committee to oversee AI implementation, ensuring patient safety and ethical use. “AI will not replace doctors but empower them to provide more personal, efficient care,” Nir emphasizes.
Opening the Ecosystem to AI Innovation
Tzafrir Kagan, CEO of Elad Health and of Chameleon, calls the migration a strategic shift. “Our goal is to optimize the caregiver-patient encounter through technology that delivers the right information at the right moment,” he says.
According to Kagan, the change is not only technical but cultural: “We must be AI-first. That means developing AI solutions in-house and enabling startups and partners to connect to our system. To this end, we built Layer X, which integrates securely with Chameleon EMR under strict certification.”
He points to efficiency gains: “Anesthesiology prep can take hours. With AI tools, it takes minutes. The ROI is enormous, less administration, more patient care.”
Looking ahead, he envisions even broader applications: “We are working on personalized antibiotics and real-time checks for drug interactions. All of this becomes possible once the system is in the cloud and open to advanced AI capabilities.”