In my second year of college, I was selected as a research intern at the Centre for Civil Society — the only undergraduate, I was told, from a couple of hundred applicants, seated alongside nine postgraduate researchers. The project was part of UNESCO's Global Education Monitoring work: a study of the regulation of budget private schools across South Asia. My job was unglamorous and perfect — document and analyze profiles of over five hundred low-fee private schools across India, and present what I found.
If you grew up outside India, "private school" probably means privilege. Here, it often means a three-room building in a working-class neighborhood charging a few hundred rupees a month, run by a proprietor who is simultaneously the principal, the accountant, and the person who fixes the fan. Tens of millions of Indian children attend schools like these. Their parents — drivers, tailors, vegetable sellers — pay fees they can barely afford because they have judged, customer by customer, that the free government school down the road is not worth its price of zero. [1]
Regulated like Eton, funded like nothing
What the profiles showed, school after school, was a sector caught in a regulatory uncanny valley. The rules — land requirements, playground sizes, teacher salary mandates — were written with elite institutions in mind, so budget schools survive in a permanent state of technical illegality, vulnerable to closure and, therefore, to rent-seeking. Meanwhile no serious public system measured the thing parents actually cared about: whether children were learning. The state tracked inputs — rooms, ratios, registers — because inputs are easy to inspect. Learning outcomes sat unmeasured on what I started calling the other ledger: the one where the numbers that matter most are kept, by nobody.
Tech companies measure everything that matters to them and act on it within the week. The social sector matters more and often measures less — not for lack of caring, but because nobody built the pipes.
I've spent the years since inside consumer tech, and the contrast has never stopped being instructive. At Meesho and Dashverse, if a funnel drops two percent, an alert fires, a dashboard reddens, a meeting appears on someone's calendar by Thursday. A child can pass through five years of schooling without learning to read a paragraph, and in most systems no alert fires anywhere. The feedback loop isn't slow. For practical purposes, it doesn't exist.
Why I keep this ledger open
This is the honest explanation for the two years I spent attempting the civil services, and for the shape of whatever comes after product management. It is not that I think tech is trivial — building things twenty million people use has taught me more about execution than any policy paper could. It's that the skills consumer tech beat into me — instrumentation, experiment design, cohort thinking, the discipline of writing down what counts before counting it — are nearly absent in the places where the stakes are children's lives rather than session frequency. [2]
Somewhere in my files there are still five hundred school profiles: enrollment numbers, fee structures, the occasional note in the margin about a proprietor teaching three subjects because a teacher had quit. They are the reason this site's name is not a pun to me. The grey area between the state and the market, between what we measure and what we merely regulate, between the ledger we keep and the one we owe — that's not where I got lost on the way to a career. It's where I'm eventually headed. The product work, as much as I love it, is the training montage.