One · The observation
I spent ten years at StartNet working with several hundred founders and MSMEs across Tamil Nadu, and one pattern kept appearing with an almost irritating regularity: almost nobody who succeeded started from a blank page.
Apple adapted Xerox’s interface. Flipkart adapted Amazon’s. The founders who insisted on total originality were, disproportionately, the ones who ran out of money — not because originality is bad, but because they paid for it at the wrong point in the sequence.
So I wrote it down as three moves in order. Copy a model that has already been proven, and study it properly rather than admiring it. Customize it to the market you actually operate in — language, payment, logistics, regulation, culture. Innovate on top of the customised base, where innovation is finally cheap because the foundation is already de-risked.
CoirGarden did ₹8 crore in online sales in under two years applying it.[B]Grade B · ClaimedReported by the company. Regraded from a flat claim pending sight of the underlying sales record. Listed in the Evidence Log.As at 26 August 2026 That is when I stopped calling it an observation and started calling it a framework, and Copy, Customize, Innovate became the first book.
Two · The escalation
A framework that worked for a startup did not answer the question a ₹2,000-crore group was asking. They were not trying to enter a market. They were trying to work out why an organisation with every advantage moved so slowly, and whether that was fixable or simply what scale does to you.
The answer arrived from an unlikely direction. I was reading NVIDIA’s investor presentation — not looking for a book — and the phrase that stopped me was parallel processing. AI had not merely outpaced Moore’s Law; it had changed the shape of the curve. Which raised the obvious question nobody was asking: if we know exactly which architectural properties let a machine improve a million-fold in a decade, why are we not building organisations with the same properties?
That produced two books rather than one. The Accelerated Organization took twelve of those properties and scored real organisations against them on five maturity levels. The AI Flywheel took the four AI stages that everyone was treating as a menu and showed that they are a powertrain — each one driving the next, with DRIVE as the hub and CORE as the rim.
Then goal1, because both of those books kept running into the same wall: an organisation that cannot read its own numbers truthfully cannot tell whether any of it worked.
Three · The correction
The fourth movement is the one that changed how I work rather than what I write about.
I went looking for the paperwork behind the most-repeated founder story of the last twenty years — filings, court records, loan agreements, a prospectus. Almost none of the order was right. The rocket company came first. The car company he did not found. The solar company was his cousins’. Twenty-four Decembers, and one of them very nearly ended all of it.
What was left after the myth came off was not a personality. It was a structure — and the same arrangement turned out to be sitting inside a Danish state utility, a Dutch machine shop and a carpet company in Georgia, none of them with a founder worth writing a biography about. That became Milk, Mountain, Moonshot.
But it also produced something I did not expect to need: an evidence standard I now hold myself to. Every figure gets a grade — [A] verified, [B] claimed, [C] estimate — and an as-at date. If a number cannot be graded it does not get softened. It gets removed.
A book that opens by dismantling a smoothed biography cannot carry one. That obligation runs to the author before it runs to anyone else.
It costs nothing and almost nobody does it, which is precisely why a family-office CFO notices it inside ninety seconds.
Four · What I do now
Three layers, running at the same time.
Milk — StartNet, elan1 and Publytics. The businesses that pay, that supply the datasets and the reference clients, and that make the frameworks testable on live organisations rather than on slides.
Mountain — Think TN Foundation. An independent Tamil Nadu policy institution on a long clock, with published governance and its own corrections log. It is deliberately kept at arm’s length from this site; its independence is easier to defend that way.
Moonshot — one objective at the far end, which everything above is subordinated to. Under my own standard that sentence must be physically specific and unchanged for a decade, so it is published on the Portfolio page or it is not published at all.
I am running the framework I sell. You can check it.