Context before autonomy.

I wrote my first program in the 9th grade, and shipped software for a living before anything else. Then I put the editor away — for over a decade — and went collecting the skills around the code.

First, the numbers. As a business analyst with a coder's background — Mu Sigma, then IMS in Germany — I learned how a business actually thinks: what it measures, what it fears, which numbers move and which just decorate a dashboard. And somewhere inside those models came my first taste of AI.

Then the products, carrying that habit with me. India's first keyless entry at Zoomcar. The delivery backend at early Swiggy. MIUI for India at Xiaomi — ten thousand units to market leadership. B2C, then B2B at Tredence — rebate and pricing suites that Fortune-500 CPG and manufacturing companies paid for. Every feature I shipped answered to a number — the analyst never left the room.

On June 4, 2019, an appraisal-day frustration became a pub conversation that ended with "what if we started something of our own?" Pangea Tech was the answer — bootstrapped, three founders in a small Bangalore room, built on a thesis the market took five years to catch up to: enterprise AI must be explainable, governed, and usable inside real workflows — not demonstrated in proofs of concept. I ran it as CEO for six years. The deployments above are that thesis, shipped.

And then AI closed the circle. In 2024, building with agents put my hands back on the keyboard — except the coder came back carrying the analyst's numbers, the product judgment, and the CEO's scar tissue. Shipping in weekends what teams once scoped in quarters. Every agent on this site is mine.

That's the whole pitch. Most advisors stopped building long ago; most builders have never sat in the CEO chair. I've walked the full circle — recently. Advise like an executive. Build like an engineer. Fractional, so you get all of it without the headcount.