How I got from a rural health centre in Tamil Nadu to building clinical AI evaluations — year by year, most recent first.
Ashoka University · CRASH Lab
I moved full-time into clinical AI research at the Koita Centre for Digital Health, Ashoka University — first as a Faculty Fellow, then as a Simons Ashoka Early Career Fellow from November 2025. In May 2025 I founded CRASH Lab, a clinician-led group I lead.
Our work is about knowing when AI can be trusted with a clinical decision and when it cannot. Radiology's Last Exam (RadLE) benchmarks frontier multimodal models against radiologists; the Gates Foundation-supported PrAImaan and V2DD programmes evaluate health-AI systems and multilingual AI scribes in Indian workflows; a DST–A*STAR India–Singapore programme builds equitable radiology foundation models; and PREDICT-AI, with Tata Memorial Hospital, develops oncology decision support. See selected work →
National President, FAIMA
I served as National President of the Federation of All India Medical Associations — an elected, voluntary role — after two years as General Secretary, working on resident-doctor welfare, workplace safety and medical-education policy.
Medical AI Bootcamp, Rajpurkar Lab (Harvard)
I took part in the Medical AI Bootcamp at the Rajpurkar Lab, Harvard University. It led to my contribution to MedVersa, a generalist medical-imaging foundation model, where I led the radiologist evaluation — later published in NEJM AI.
Locum Radiologist
Between residency and Ashoka, I worked as a locum radiologist at Medanta and Fortis Hospitals in Delhi NCR — a first look at private-sector reporting workflows after years in the public system.
Senior Resident, AIIMS New Delhi
High-volume emergency, inpatient and outpatient radiology — interpreting 200+ studies independently during emergency shifts, predominantly CT and ultrasonography. Alongside the clinical work, AI research became serious: my rib-fracture pipeline AIRib won the RSNA Trainee Research Prize in 2023, and I presented cutting-edge oral work on LLM-augmented speech recognition for reporting at RSNA.
M.D., Radiodiagnosis & IR, AIIMS New Delhi
Postgraduate training in diagnostic and interventional radiology. My thesis, in collaboration with IIT Jodhpur, developed automated detection, characterisation and prognostication of rib fractures on CT — the beginning of AIRib. I worked through roughly a year of COVID duties and the institution's transition from semi-digital to fully digital imaging.
Internship (CRRI), JIPMER
Rotations across core specialties, including a two-month residential posting at a rural primary health centre in Ramanathapuram — alternate emergency coverage, NCD clinics and camps, and referral decisions under real medicine and infrastructure constraints, working with patients in Tamil and Hindi. This is where my interest in how care is actually delivered took root.
M.B.B.S. (Gold Medallist), JIPMER
Beyond the degree, I coordinated CASE, a low-cost arterial-stiffness estimator selected among the top 40 innovations nationally in the DST–Lockheed Martin–Tata Trusts IIGP 2.0 challenge, and helped run Connaissance, JIPMER's international undergraduate medical conference.