QUICK TAKE | Doctors Speak, AI Types: Seven Things to Know About BHASHINI’s Prescription Tool

1. What is BHASHINI?
BHASHINI is India’s government-backed language AI platform under the Ministry of Electronics and Information Technology. It provides speech recognition, translation and other language services that organisations can integrate into existing digital systems. Its healthcare applications aim to reduce documentation work and improve communication across languages.
2. How does its prescription tool work?
At AIIMS New Delhi, BHASHINI’s speech-to-text facility connects with the National Informatics Centre’s eHospital prescription module. Doctors dictate their prescriptions, and the system converts their speech into electronic text. The announced application supports recording the doctor’s instructions; it does not establish autonomous diagnosis or selection of medicines.
3. What has changed at AIIMS?
A pilot in Neurology and Neurosurgery is moving to an institute-wide rollout. Professor Vivek Tandon, who oversees AIIMS’ computer facility, reported positive feedback from doctors. The announcement does not specify a completion date or how many departments are already using the tool.
4. Is it only for prescriptions?
JIPMER has integrated BHASHINI into its e-Sushrut hospital management system for dictating case histories, inpatient notes and discharge instructions. The Army demonstrated spoken clinical notes becoming typed records through its Dhanwantri system. The technology therefore addresses documentation across different stages of patient care.
The Wealth Issue 2026
06 Nov 2026 - Vol 05 | Issue 45
Portraits of Dreamers and Doers under 35
5. Can doctors use Indian languages?
JIPMER’s integration supports English and supported Indian languages. The exact language coverage depends on the application. BHASHINI’s wider healthcare work also includes translation and doctor-patient communication, potentially helping patients understand medical information in a familiar language.
6. Will it actually save consultation time?
Reducing typing and paperwork is the stated objective. However, the announcement provides no measured time savings, transcription-error rates or comparison of consultation lengths. Positive pilot feedback supports expansion, but published performance data would help establish how much the tool improves everyday clinical work.
7. What needs particular care in a dictated prescription?
Medicine names, strengths, doses, frequency and duration must be recorded accurately. A transcription error in any of these fields could change the instruction. Clinical review of the resulting text is therefore essential. The announcement does not detail its verification workflow, patient-data safeguards or procedures for correcting errors.
