From US Residency Programs to EdMedAI — Why I Built India's First AI-Powered CBME Platform

I did not set out to build a technology platform. I set out to solve a problem that I had watched grow — quietly, stubbornly — for more than two decades.

I am a medical education executive. I have spent over 25 years working inside the graduate medical education system of the United States — building programmes, training doctors, and chairing national medical education committees. I have seen, from the inside, what a functioning competency-based training system looks like. And I have seen, from a distance, what happens when one is mandated on paper but never truly built.

That gap is why EdMedAI exists.

Starting Residency Programmes from Zero

Early in my career in the United States, I was involved in something most medical educators never experience: starting a graduate medical residency programme from nothing. Not inheriting an established programme. Not tweaking an existing curriculum. Starting from a blank sheet — designing the training structure, building the faculty base, getting accreditation, and then recruiting the first class of residents.

I did this more than once. I led the conversion of teaching hospitals into accredited graduate medical education institutions and established residency programmes — in Internal Medicine and in Psychiatry — that went from concept to accredited operation. I also worked to grow established programmes, increasing the number of positions and expanding the scope of training to bring in more International Medical Graduates and American Medical Graduates alike.

What that experience taught me is something that sounds simple but is actually profound: the quality of a doctor is not a product of how talented they were when they entered training. It is a product of the system they trained inside. A well-structured programme with clear milestones, consistent faculty mentorship, and regular competency assessment produces better doctors — reliably, predictably — regardless of where those doctors started.

This is not a theory. It is what the evidence from decades of outcomes data in American graduate medical education consistently shows. And it is what I observed directly, watching residents develop over three and four years inside programmes I had helped build.

Leading Co-Management Programmes — Quality, Safety, and Patient Outcomes

Alongside my work in residency training, I led physician co-management programmes at American hospitals — a model where physicians from different specialties collaborate in a structured, data-driven way to manage complex patients who fall across traditional department boundaries.

The results were measurable and significant. Across the hospitals where I led these programmes, we achieved:

• Reduced care variance — patients with the same condition receiving more consistent, guideline-aligned treatment regardless of which team was on duty
• Improved patient safety — measurable reduction in adverse events and near-misses
• Reduced clinical waste — eliminating redundant or non-evidence-based investigations
• Increased EMR utilisation — clinical data actually being captured and used for decisions, not filed away
• Reduced cost of care — systemic savings from better coordination and appropriateness review
• Increased patient satisfaction — more coordinated, communicative care experiences

All of it pointed in the same direction: better patient outcomes.

What this experience demonstrated — and what I brought directly into EdMedAI's design — is that quality improvement in healthcare is not a matter of individual physician effort. It is a matter of system design. When you build the right structure around clinical teams — clear protocols, shared data, regular audit, and accountability loops — quality improves predictably. When you leave it to individual initiative without structure, you get variance. And in medicine, variance kills.

This is exactly the problem I saw in Indian medical education. Not a shortage of talented doctors or dedicated faculty. A shortage of structure — the data systems, the feedback loops, the accountability mechanisms — that transforms individual talent into consistent, measurable excellence.

Teaching Value-Based Care, Medical Economics, and Digital Health

One of the responsibilities I took most seriously as a medical educator was preparing residents for the healthcare system they were actually going to practise in — not just the clinical one, but the economic and systemic one.

I taught residents about medical economics: how healthcare spending works in the United States, how individual clinical decisions aggregate into systemic costs, and how physicians can be stewards of resources without compromising patient care. The core lesson was straightforward but often absent from medical education: ordering a test that will not change your management is not thorough medicine — it is clinical waste. Evidence-based medicine is not just about efficacy. It is also about appropriateness.

I taught co-management principles and value-based care — the shift in American healthcare from fee-for-service models to outcomes-based payment, and what that demands from physicians in terms of documentation, quality tracking, and population health thinking. I taught physician leadership — how doctors can and must take active roles in the institutional and policy decisions that shape how care is delivered.

On digital health, I went beyond teaching — I implemented it. In hospitals and clinics across the United States, I led the adoption of digital health tools: electronic medical record systems, clinical decision support, data analytics platforms, and telehealth infrastructure. Getting a hospital or clinic to genuinely use a digital health system — not just have it installed, but have physicians actually entering data, reviewing alerts, and acting on analytics — requires change management as much as technology. I led that change in real clinical environments, which gave me a direct understanding of where digital health implementation succeeds, where it fails, and why.

Beyond residency training, I had the opportunity to bring these ideas to academic audiences. I was invited as a guest lecturer at Boise State University, where I spoke on price transparency and value-based care — two of the defining challenges reshaping American healthcare and, increasingly, healthcare systems worldwide.

I also conducted AI lectures for medical and dental students, physicians, and other learners — introducing them to how artificial intelligence is already reshaping diagnostics, clinical decision support, radiology, pathology, and the day-to-day workflow of a practising physician. My message to every cohort was the same: AI will not replace doctors. But doctors who know how to work with AI will replace those who do not. That is not a distant future scenario — it is happening now, in every well-resourced hospital system in the world. Medical students, dental students, and practising physicians who have not developed AI literacy are entering a profession that has already moved on without them.

International Medical Graduates — What I Learned

A significant part of my work in American residency programmes involved International Medical Graduates — doctors who had completed their medical degrees outside the United States and were seeking residency training to practise in the US. India is consistently one of the top sources of IMGs who match into American residency programmes. Many of the most talented residents I worked with over the years had trained in Indian medical colleges.

What I noticed — and what shaped much of my thinking — was a striking gap between their raw clinical intelligence and their structured training documentation. Indian medical graduates who matched into US residency programmes were, almost universally, highly motivated, deeply knowledgeable, and exceptionally hardworking. But they often arrived having been assessed informally, without structured competency evidence, without clear milestone documentation, and without the habit of systematic clinical reasoning documentation that American medical education demands from day one.

This was not their failure. It was their system's gap.

The American residency programme had to accelerate these residents' adaptation to a documentation-heavy, milestone-driven, faculty-observed environment. Most of them adapted quickly and went on to exceptional careers. But the adaptation took time — time that would not have been necessary if their undergraduate medical training had already embedded the habits of competency documentation, structured reflection, and systematic clinical reasoning.

I also worked with American Medical Graduates throughout my career — and the difference that structured CBME training made for them was equally clear. AMGs who came from programmes with rigorous milestone tracking, regular direct observation, and strong formative feedback arrived at residency ready to be assessed. They were not starting from zero. The system had already taught them how to learn — not just what to learn.

Chairing the Graduate Medical Education Committee

Serving as Chairman of the Graduate Medical Education Committee in the United States gave me a perspective on medical education that goes beyond any single programme or institution. In that role, I was involved in policy — how standards are set, how compliance is measured, how accreditation decisions are made, and what separates the programmes that consistently produce excellent doctors from those that struggle.

The answer is almost always the same: structure. Programmes that produce the best doctors are not the ones with the most prestigious faculty titles or the most advanced hospital technology. They are the ones with the most consistent structure — clear expectations, regular assessment, reliable feedback loops, and faculty who are genuinely invested in resident development.

What the accreditation process also showed me is how powerful data is when it is collected consistently. Programmes that track outcomes — resident progression, assessment scores, clinical exposure counts, formative feedback quality — can identify and correct problems early. Programmes that do not track systematically discover problems only when they become crises: a resident who has progressed to year three without adequate skill development, or a cohort that struggles on board examinations because a gap in curriculum was never caught and corrected.

These insights shaped every design decision I made when building EdMedAI — especially the analytics dashboards for HODs and Principals, the mandatory faculty verification on logbook entries, and the real-time NMC compliance tracking.

Building AI-Native Healthcare and Medical Education Products

Everything I had done in residency training, co-management, value-based care, digital health implementation, and teaching AI literacy to clinicians converged into a natural next step: building AI-native products myself.

I developed multiple AI-native products spanning both medical education and healthcare — tools that did not simply layer AI features onto existing software, but were designed from the ground up around AI capabilities. These products applied machine learning, clinical decision support intelligence, and natural language processing to real problems in medical training and clinical workflow.

The most significant of these was a multi-agentic AI platform — a system in which multiple specialised AI agents work in coordination, each handling a distinct function, with outputs feeding into one another to produce results that no single AI model could achieve alone. In a healthcare and medical education context, this matters enormously:

• A single AI model can generate a quiz question, or summarise clinical evidence, or flag a compliance gap
• A multi-agentic system can do all three simultaneously, cross-check each output against the others, adapt based on context, and produce a coherent, audited result
• At scale — across thousands of users — without degrading in quality

This is the architecture that underpins EdMedAI. When EdMedAI generates a case study grounded in a specific NMC competency, evaluates a student's quiz performance to adjust their spaced repetition schedule, and simultaneously flags a faculty member's logbook attestation queue — all in a single session — that is a multi-agentic system at work. The experience of building that infrastructure in the United States, under real clinical and educational performance requirements, is what made it possible to build EdMedAI at the depth and scale that it operates.

Healthcare Leadership, Regulatory Experience, and National Advocacy

Running programmes and building products is only part of what shapes a healthcare executive. The other part is operating inside the regulatory and accreditation frameworks that govern American healthcare — not as an observer, but as a participant.

I have been part of accreditation site visits and Joint Commission accreditation visits, as well as other US healthcare regulatory processes. These experiences give you a particular kind of clarity: you see exactly what separates institutions that are genuinely compliant from those that are compliant only on paper. The difference is almost always documentation — the quality, consistency, and retrievability of records that demonstrate that standards are actually being met in daily practice, not just on inspection day. That insight is embedded in every compliance and audit feature EdMedAI has built.

On price transparency, I did not just follow the US federal mandate — I championed it. When the Centers for Medicare and Medicaid Services mandated hospital price transparency, I built products to help hospitals become compliant and educated the stakeholders who needed to act: hospital administrators, finance teams, physicians, and policymakers. I conducted multiple webinars on price transparency, translating a complex regulatory requirement into practical steps that healthcare organisations could implement.

Across my career in the United States, I have given more than 100 presentations to non-profit and for-profit healthcare organisations nationwide — covering medical education, AI in healthcare, value-based care, digital health, and physician leadership. These ranged from small departmental briefings to large national conferences, reinforcing a discipline that is directly relevant to EdMedAI: the ability to take complex ideas and translate them into language that moves people to act.

I have also served on the American College of Healthcare Executives (ACHE) Board, Alabama Chapter — a recognition of peer standing within the US healthcare leadership community and an opportunity to shape the development of healthcare executives across the state.

What I Saw When I Looked at India

India has some of the most talented medical minds in the world. Indian doctors are in leadership positions across American hospitals, British NHS trusts, Australian health systems, and every major global medical institution. The raw material is exceptional.

But the system inside which Indian undergraduates train has, for decades, been built around a fundamentally different model: volume over structure, memorisation over competency, summative examination over longitudinal development. The NMC recognised this clearly when it mandated Competency-Based Medical Education in 2019. The CBME framework is genuinely well-designed — the 2,683 competencies, the K/KH/SH/P levels, the DOAP pedagogy, the AETCOM curriculum, the mandatory logbook — all of it reflects serious thinking about what a medical graduate should be able to do.

The problem I saw was implementation. Not intent — execution.

When I visited medical colleges and spoke with faculty and principals, I found the same pattern everywhere: CBME was being documented, but not truly delivered. Logbooks were being filled, often retrospectively. DOAP sessions were being recorded, frequently without genuine structured faculty observation. Faculty — already stretched thin across large student cohorts, heavy clinical loads, and administrative demands — had no practical tools to do any of this systematically.

The gap between what the NMC had mandated and what was actually happening in wards and classrooms across 816 medical colleges was enormous. And it was widening, not closing.

The Decision to Build EdMedAI

The decision to build EdMedAI came from a very specific realisation: the tools Indian medical education needed did not exist. What existed were generic learning management systems designed for other sectors, or homegrown college databases that captured data but generated no intelligence from it.

What was needed was a platform built from the ground up around the NMC CBME framework — one that understood the difference between a K-level and a P-level competency, that could track DOAP progression across 4.5 years, that could generate AI-powered learning content grounded in specific NMC competency codes, and that could give principals, HODs, faculty, and students the information they needed to do their jobs better.

No such platform existed. So I built it.

Every feature in EdMedAI traces back to something I observed in American graduate medical education that was missing from the Indian system:

• The digital logbook with mandatory faculty verification traces back to the direct observation requirements of American residency
• The spaced repetition quiz system traces back to the evidence on how clinical knowledge is retained long-term
• The 50+ clinical simulations trace back to simulation laboratories that American programmes use to build procedural skills safely
• The AI-generated case studies trace back to the case-based learning model that American medical education has used for decades
• The HOD and Principal analytics dashboards trace back to the programme director dashboards I used to track resident progression
• The DOAP skills tracker is a direct translation of the milestone tracking system used in American residency programmes

I did not invent these ideas. I translated them — from the American graduate medical education context, where they have been proven over decades, into the Indian undergraduate medical education context, where they were urgently needed.

Why AI — and Why Now

The AI component of EdMedAI is not a feature added for marketing appeal. It solves a specific, structural problem in Indian medical education: the impossibility of providing high-quality, personalised learning support to hundreds of students simultaneously with the faculty-to-student ratios most Indian medical colleges operate under.

In American residency programmes, a resident has a programme director, multiple attending physicians, and a structured feedback system — all focused on their individual development. In Indian medical colleges, a faculty member may be responsible for 60, 80, or even 100 students simultaneously. Individual feedback, personalised content, adaptive assessment — these are structurally impossible at that ratio without technology.

AI changes the equation. An AI tutor that generates case studies grounded in a student's specific NMC competencies, an AI question bank with 80,000+ adaptive MCQs mapped to every competency, an AI assistant that answers CBME curriculum questions at any hour — these are not replacements for faculty. They are force multipliers that allow the faculty India has to teach at a quality level that was previously possible only in programmes with much more favourable student-to-faculty ratios.

What I Want for Indian Medical Graduates

The doctors India trains today will be treating patients in 2040 and beyond. The medical knowledge that will matter most then does not fully exist yet. What will always matter — what no AI will replace — is the ability to think critically in the face of uncertainty, to communicate with a frightened patient with genuine empathy, to lead a clinical team under pressure, and to keep learning across an entire career.

These are exactly the capacities that Competency-Based Medical Education is designed to develop. And they are exactly the capacities that EdMedAI is built to support.

My goal is not for EdMedAI to be used in 10 colleges or 50 colleges. My goal is for every NMC-recognised medical college in India — all 816 of them — to have access to tools that give their students the same structured, evidence-based, technology-supported learning environment that the best residency programmes in the United States provide to their residents.

Indian medical graduates are already among the best doctors in the world. With the right training system behind them from day one of MBBS, they will be in a category of their own.

That is what EdMedAI is for. That is why I built it.