Amrita Unveils MedSum, an AI Clinical Assistant that Turns Doctor-Patient Conversations into Medical Records

The medical summaries can be delivered as voice messages in several Indian languages, including Hindi, Tamil, and Bengali, allowing patients to replay instructions related to diagnosis, medication, or treatment plans.

Amrita Vishwa Vidyapeetham has unveiled MedSum, an AI-powered clinical assistant designed to convert doctor-patient conversations into structured clinical documentation and multilingual summaries.

The tool aims to reduce administrative workload for clinicians while improving the clarity of medical information delivered to patients.

The system has been developed by the School of Artificial Intelligence at Amrita Vishwa Vidyapeetham’s Faridabad campus in collaboration with Amrita Hospitals and Amrita Technologies.

During consultations, MedSum listens to the discussion between the doctor and the patient in real time and generates structured medical records, prescriptions, and patient summaries through a clinician-facing dashboard.

Beyond documentation, the platform focuses on improving patient understanding of medical advice. After a consultation, MedSum can produce simplified summaries of the discussion in the patient’s preferred language.

These summaries can also be delivered as voice messages in several Indian languages, including Hindi, Tamil, and Bengali, allowing patients to replay instructions related to diagnosis, medication, or treatment plans.

This feature is particularly useful for elderly patients or individuals who may not be comfortable with English-language medical records.

“Documentation can consume a significant part of a doctor’s day and often impacts the quality of patient interaction. MedSum helps us focus more on clinical care while ensuring accurate records,” said Dr Sachin Gupta, Dermatologist, Amrita Hospital, Faridabad.

MedSum is already being used across several departments at Amrita Hospitals, including gynaecology, oncology, neurosurgery, and paediatric neurology. The system operates through a combination of tools that include a doctor dashboard, a patient application, and an administrative console that allows hospitals to manage documentation workflows.

Prof. Kamal Bijlani, Dean of the School of Artificial Intelligence at Amrita Vishwa Vidyapeetham, said the system was built with Indian healthcare requirements in mind. “MedSum runs on BharatGen, India’s own large language model, and has been validated in real clinical settings. Our focus has been to deliver reliable documentation and accessible patient communication at scale,” he said.

The platform operates on the Param-17B BharatGen model, part of India’s family of indigenous large language models designed for Indian languages and contexts. By relying on domestic AI infrastructure, the system represents one of the early healthcare applications built entirely on India’s sovereign AI ecosystem.

Stay tuned for more such updates on Digital Health News