About
I started in theoretical physics. My PhD was in quantum chaos, which taught me to sit with a hard system long enough to see its structure, and then the rigor to abstract it into a model that holds. I have spent the twenty-five years since applying that rigor where being wrong is expensive: fraud and credit models at global banks, anti-money-laundering systems, energy analytics for Fortune 500 utilities, real-time pricing across a hundred-plus cities at a global ride platform, telecom, and most recently health, at Resmed, where I built the AI function from scratch and shipped it into FDA-cleared products used by millions.
The work changed as I rose. Early on I built the models myself. Over time, the decisions that determined whether they shipped moved up a level: which problems to go after, who to hire and how to build the team, how to bring product, clinical, regulatory, and finance onto one plan, and how to set a strategy a board would fund. By the end I was helping set AI strategy for Resmed, and building both the technical foundation and the organization that carried it.
Production AI is two hard problems, not one.
The lesson that kept repeating across that arc: production AI is two hard problems, not one. Getting the model right, from architecture through validation, bias, guardrails, and drift, is hard and takes years to learn. Getting it into a regulated product, a real workflow, and a P&L, and governing it so it stays trustworthy, is just as hard. The five pillars I hold every engagement against, product, productivity, data foundations, governance and trust, talent and organization, codify the decisions that arc forced at every level. I can speak to each pillar because I have owned each one, first as the engineer building it, later as the executive answerable for it.
Today I advise a Fortune 100 healthcare enterprise through this practice, serve as a Contributing Analyst at GAI Insights researching AI for private-equity and Fortune 100 clients, sit on the selection committee at the EvoNexus incubator, and serve on the Consumer Technology Association's Health AI Planning Council. I write about applied AI, governance, and the economics of getting it into production on Substack.
Education
Badri Raghavan, Ph.D. · Theoretical Physics, Northeastern · MS Physics, Carnegie Mellon · MSc Physics, IIT Kanpur · BSc Honours Physics, St. Stephen's College, Delhi
Advisory
CTA Health AI Planning Council · CurieAI, Board Advisor · my3dmeta, Board Advisor · EvoNexus Selection Committee · GAI Insights, Contributing Analyst