I had the range early. At Tata Sons' investment office I sat on the buy side across a wide set of businesses: airlines, hospitals, hotels, power, infrastructure, auto, precision contract manufacturing. As the shareholder, you see both halves: the investing case and how it actually plays out inside the company. Seeing that many models up close, that early, builds a kind of pattern recognition,  the surface of every business is different, but only a handful of levers decide whether it works. You stop seeing sectors and start seeing structure.

‍

Then I went deep, through the underwriting discipline of private equity board seats, large cheques behind real diligence. Breadth gave me the patterns; depth turned them into a framework. What I find most alive now is the exceptions,  the company that wins on a lever the framework said wouldn't matter. A view held firmly enough to act on, and loosely enough to keep learning.

Investing is the art of underwriting risk. I'm not looking for the deal with no flaws:  that one's either imaginary or overpriced. I'm looking for the risks I understand well enough to take on purpose, and to carry. That's the craft: choosing which things to be wrong about. And then luck gets a vote. The same honest process can meet a good outcome or a bad one, and you rarely get to fully claim either — which is what keeps me humble, and a little stoic.

Financial services is my highest-conviction sector, and it comes down to reach. It's the one sector whose value reaches all 1.4 billion people, not just a slice at the top. The infrastructure to serve them has been in place for a while, but the economics of going that deep never quite worked. 

‍

What's changed is AI: it takes two to three points out of opex and makes serving deeper Bharat viable for the first time. The regulator has enabled it, and distribution, the one thing AI can sharpen but never replace, only grows more valuable: lending, insurance, wealth, broking. That's the rare setup: a sector that gains from the AI tailwind instead of being disrupted by it. Returns you can see, not a story you tell.

AI is the wildcard, and as it lands, profit pools shift across every industry. Winners come from reading where the pool is actually moving, not where the story is loudest. Three places I want to double down. 
‍

1. Financial services, for the structural reasons I'd defend anywhere. 

2. AI and services, where India stands to gain the most: we have the cheapest human-in-the-loop cost in the world, and when AI automates 60 to 80% of a task, new companies enter on price and domain depth, taking share from incumbents who can't restructure as fast. 

3. Manufacturing, as AI moves into everyday physical devices and productivity finally has to be felt off the screen

The misallocation is simple: capital is chasing the surface of the AI narrative instead of the structure beneath it.

Because I see businesses through their structure, I'm looking for a founder who sees their own that way too — and then shows me something I'd missed. The best ones understand their business beneath the surface: they know what truly drives it, they don't look away from the hard parts, and they read the shape of their market more honestly than most. They don't argue with reality; they move with it, often faster than I can. The conviction I'm chasing is when the founder and the framework arrive at the same truth from opposite directions — mine from outside, theirs from within.