ByDR. ITAY GAL

AUGUST 5, 2026 14:35

Assuta Medical Centers and Microsoft are building an artificial intelligence platform that functions as a dedicated helper for physicians. The tool gathers pertinent health data, condenses test outcomes and prior notes, records patient encounters, and generates draft summaries automatically.

Rather than substituting for the physician, the AI reduces time spent on paperwork and information retrieval, freeing more hours for direct exams, dialogue, and clinical choices.

Unlike conventional electronic health records that require physicians to toggle between multiple screens, the AI agent reviews the full record and surfaces essential points before each appointment.

Prior to the patient arriving, the physician receives a concise overview of past diagnoses, current medications, recent test results, prior treatments, and any shifts in condition. This approach cuts down on time spent hunting through extensive files and lowers the chance of missing key details.

Once consent is secured and privacy standards are met, the system records the doctor-patient exchange during the visit and creates a structured transcript. It then compiles a full clinical note covering complaints, history, findings, assessment, and follow-up suggestions.

The physician must examine, edit, and authorize the note before it enters the official record. Assuta emphasizes that the platform serves only as support and never replaces clinical judgment; final responsibility for diagnosis and care stays with the doctor.

After the visit, the system can flag needed tests or follow-up appointments, identify open slots, and assist with scheduling based on clinical priority and system availability. Further features may be added later under medical supervision and data-protection rules.

AI platform aims to trim administrative load while preserving doctor-patient interaction time

A major strain on health systems is the hours physicians devote to documentation, typing summaries, and administrative duties, often extending past clinic hours. Project goals include saving hundreds of physician hours monthly through the agent.

These efficiencies could shorten overall visit length without cutting conversation time between doctor and patient.

Instead of typing during much of the appointment, the physician can let the system handle documentation, potentially allowing more patients per day, greater appointment access, and shorter clinic wait times.

Building the system goes beyond linking transcription software. The agent must train extensively on medical documents, summaries, test results, and terminology while stripping identifiers to safeguard privacy.

Training focuses on separating critical facts from minor ones, linking diagnoses with medications and tests, and producing tight summaries that retain vital information. Medical staff review outputs, note mistakes, and refine how data is presented for clinical use.

A key hurdle is avoiding typical AI pitfalls such as dropped details, patient mix-ups, or invented content. Therefore, no output enters the record without human review. Rollout will proceed step by step, testing accuracy, safety, and impact on team workflows.

The initial pilot is slated to start within a year on a limited basis, allowing Assuta to assess integration, time savings, summary quality, and comfort levels for both staff and patients with automated note-taking.

Microsoft plans additional AI agents for healthcare

Beside the personal assistant, Assuta and Microsoft will move to the company's cloud platform and build an enterprise system for creating and running more AI agents. The setup may later support tools in clinical care, customer service, and operations while upholding strict privacy and security standards.

Sarit Sass, healthcare organizations director at Microsoft Israel, noted that the future of healthcare depends on blending medical team expertise with artificial intelligence capabilities, with the partnership intended to let clinicians concentrate on patient care.

Ofir Shani Shaharbani, head of Assuta's Information Systems and Technology Division, called the collaboration a major step forward that will embed AI responsibly and securely at the core of medical work while protecting patient privacy.