What Lies Beneath

The Weight of Foundations

There is a peculiar truth about buildings that most people only notice when the ceiling leaks: the part that matters most is the part you never see. The foundation. The plumbing. The load-bearing walls disguised as ordinary plaster. We celebrate the glass façade and the rooftop view, but the structure’s real argument with gravity happens underground, in silence, in concrete.

This is not merely an architectural observation. It is a philosophy of work, and it is the only honest way to understand how I came to spend two decades thinking about healthcare data.

My path did not begin with a corporate master plan. It began in Calcutta, in the dense, argumentative, intellectually voracious ecosystem of a traditional middle-class family where education was not an extracurricular activity but the primary engine of social mobility. In the competitive crucible of English-medium schooling, one learned early that clarity of thought and precision of language were not polite accomplishments; they were instruments of leverage. While the world around me offered its comforts and distractions, I found my bearings elsewhere—in books, in the austere satisfactions of foundational mathematics, and in a persistent, almost tactile curiosity about how complex systems hold themselves together.

That curiosity is a dangerous thing. It does not let you rest with surfaces. It compels you to ask what lies beneath the interface, beneath the protocol, beneath the presentation. It is the instinct that makes an architect out of a programmer, and a skeptic out of an enthusiast.

One Suitcase and a Carry On

In August 1998, I crossed an ocean with a single suitcase, reserves that could be described charitably as modest, and a determination that could be described accurately as stubborn. The destination was the United States, and the objective was not merely a degree in computer science but an immersion in the deeper grammar of technical systems—the logic, the constraints, and the often unspoken assumptions embedded in software architecture.

What followed was a twenty-year apprenticeship in one of the most demanding classrooms available: the architectural core of American healthcare information technology. This is not a field for the faint of heart. Healthcare data is not like other data. It is messy, consequential, and encoded with the full weight of human vulnerability. A misplaced decimal in a financial ledger costs money. A misplaced semantic mapping in a clinical record can cost something far more serious.

I spent these years inside massive federal data frameworks—most notably the Veterans Affairs systems—where the scale of the data was matched only by the complexity of its provenance. My work often involved the foundational modernization of medical records: the high-stakes, painstaking migration of critical clinical data from legacy hierarchical databases like MUMPS (a venerable but idiosyncratic system that powers much of American healthcare infrastructure) into robust, relational SQL environments.

This was not a technical exercise in the abstract. It was archaeology. It was translation. It was diplomacy. You are not merely moving bits from one format to another; you are deciphering decades of undocumented clinical intent, institutional habit, and silent assumptions baked into data structures that predate the modern internet. You learn, very quickly, that interoperability—the ability of different systems to exchange and make sense of information—is not a checkbox feature. It is a form of semantic liquidity. It requires that data not only travel, but arrive with its meaning intact.

The Clinician’s Revenge

There is a particular kind of hubris common in software engineering, and I encountered it repeatedly during those years. It is the belief that a system designed elegantly in a vacuum will necessarily function elegantly in the wild. Early electronic health record implementations, for all their good intentions, often created more friction than they solved. The reason was not technological failure; it was anthropological failure. They were engineered by individuals who had never stood in a clinic at 2 AM, who had never watched a physician wrestle with a workflow that did not match the software’s assumptions, who had never felt the visceral urgency of a clinical decision made under time pressure.

This realization shaped my professional philosophy with the force of a correction: true digital health architecture requires an uncompromising alignment between engineering execution and clinical intent. You cannot draw the blueprint without understanding the living structure that must inhabit it. The database schema must respect the clinical narrative. The ETL pipeline must honor the physician’s cognitive workflow. The interface must serve the patient, not the other way around.

Years of working with enterprise standards like HL7 and evolving frameworks like FHIR (Fast Healthcare Interoperability Resources) reinforced this conviction. Standards are necessary but insufficient. They are the grammar of a language; the poetry is in the implementation. And the poetry is always harder than it looks.

The Return to build in a Different Key

Eventually, life completed its arc and brought me back to India. I returned not with nostalgia, but with a specific aspiration: to build meaningful digital health infrastructure and to drive national-scale interoperability in a landscape that desperately needed both. What I found was a masterclass in the politics of transformation.

Digital health in emerging markets is not a linear narrative of progress. It is a labyrinth of competing stakeholder interests, socioeconomic realities that refuse to simplify themselves, and the structural inertia that accompanies any attempt to change how institutions handle information. My entrepreneurial endeavors in this space—founding and leading ventures focused on health information exchange, clinical trial systems, and structured data platforms—were not romantic exercises in innovation. They were tests of endurance, corporate governance, and the capacity to maintain architectural integrity while navigating environments that often reward speed over soundness.

These years reinforced a belief that I hold with increasing conviction as the industry matures: successful modernization demands long-term organizational commitment over temporary technological enthusiasm. The shiny tool will be replaced by a shinier tool in eighteen months. The data architecture, if built correctly, should outlast them both.

First Principles in an Age of Noise

Today, the technology landscape is loud. It is a marketplace of marketing hype, automated shortcuts, and the relentless promise that artificial intelligence will solve problems we have not yet bothered to define properly. As the industry rushes toward deploying AI and predictive modeling in healthcare, my focus has narrowed rather than expanded. I am interested in the substrate.

Because here is the uncomfortable truth that no one puts on a billboard: AI in healthcare does not fail at the algorithmic layer. It fails at the data layer. It fails when the underlying data architecture is brittle, semantically ambiguous, or structurally incoherent. It fails when the training data carries the scars of poor interoperability, inconsistent terminology, and undocumented clinical assumptions. A predictive model is only as intelligent as the data it ingests, and most healthcare data is not yet intelligent enough to be trusted.

My work today is anchored in first-principles thinking—the disciplined refusal to accept inherited assumptions without interrogation. Before we talk about models, we must talk about meaning. Before we talk about automation, we must talk about governance. Before we talk about intelligence, we must talk about integrity.

What This Space Is For

This platform is a reflection of that ongoing commitment. It is not a chronicle of industry trends. It is not a repository of product reviews or a megaphone for the latest acronym. It is a space dedicated to looking past the superficial layers of technology and focusing on the rigorous engineering, statistical modeling, and architectural governance required to make digital systems genuinely useful in the domain where usefulness matters most: human health.

I write and design not to follow the market, but to document a persistent intellectual journey. The subject is the hard, hidden work of our field—the structural decisions that determine whether a system will endure or collapse, whether data will illuminate or mislead, whether technology will serve the clinician or merely impress the boardroom.

For anyone navigating the intersection of complex systems, technical data, and systemic transformation, this record is an invitation. Not to agree with every conclusion, but to look at architecture from the ground up. Because the foundation is where the argument with gravity happens. Everything else is just decoration.