Back

From Physics Classrooms to Product: An Unlikely Path That Made Sense

4 MINS

From Physics Classrooms to Product: An Unlikely Path That Made Sense

I didn't take a straight line to product. I taught IIT-JEE physics, ran a food company, planned inventory at a B2B marketplace, and now I'm building a supply-chain product. On paper it looks scattered. In practice, each step taught me a piece of the same craft: take a messy real-world problem and build a system that reliably solves it.

Teaching taught me to find the root cause

As Head of Physics, I was responsible for improving the performance of 60 students preparing for the hardest exam in India. You can't do that by working harder at the symptoms. A student who keeps getting the same kind of problem wrong has one missing concept underneath. Find it, fix it, and a dozen surface errors disappear.

Product is the same discipline. A pile of complaints is usually one root cause wearing different costumes. I learned to keep asking "why" until the real gap shows up, because fixing the symptom just means you'll see it again next week.

Running a business taught me unit economics

I started Kitchen Engineers and made it operationally profitable by the fifth month, growing 15% month-over-month for the first year. None of that came from a grand strategy. It came from staring at cooking costs like a microeconomics problem and cutting wastage from 5% to 2%, one process at a time.

That habit, treating every cost and every gain as a number you can move, is exactly what inventory work demands. Overstock is wastage. A stockout is a lost sale. Both are line items you can shrink if you're willing to look closely enough.

SOPs taught me that consistency is a system

The hardest part of a small food company was keeping quality consistent without raising costs. The answer wasn't heroics; it was SOPs, clear, repeatable processes that didn't depend on any one person having a good day.

I think about product the same way. A good forecast or a good replenishment plan shouldn't rely on the smartest planner being in the room. It should be a system that produces a reliable decision every time. The goal of any tool I build is to make the right call the default, not the exception.

Why it all points at EasyReplenish

Looking back, the through-line is obvious: find the root cause, respect the unit economics, build the system that makes good outcomes repeatable. Inventory optimization for apparel brands needs all three at once. That's why this path, scattered as it looks, feels less like a series of detours and more like a long apprenticeship for exactly the thing I'm building now.

Background

Varun skipped presentations and built real AI products.

Varun Ravula was part of the April 2026 cohort at Curious PM, alongside 18 other talented participants.