When Mukesh Bansal is asked what he is now, he doesn't say founder or investor. He says "a learner", and admits that the moment a company starts working, he gets bored. That restlessness is the thread through a career most people would have parked long ago: he founded Myntra and sold it to Flipkart, co-founded Cult.fit and turned fitness into a category, and now runs the enterprise-AI company Nurix, co-leads the Meraki Labs venture studio, and backs companies from Groww to Skyroot. This conversation follows the shape of all of it rather than any single company: why he reframes pivots as daily "adjustments", why he thinks the AI era rewards the hands-on operator over the people-manager, why rationally no one should start a company, and what his string of failed Silicon Valley startups taught him. It closes on Haridwar, three books, the longevity basics he trusts over biohacking, and a life he rates an 8. Chapters 0:00 Cold open: "willing to work on it for the rest of our lives" 0:11 Introduction 4:21 One studio, two incubations, and Cult.fit: where the time goes 5:23 "I see myself as a learner" (and why he gets bored) 7:36 When does learning stop? Crisis as a gift 11:25 Why an hour of his time isn't what it was 11:41 The three kinds of professionals in the AI age 15:30 His stack: Warp, Claude, eight to ten agents overnight 18:03 "If you're only a people manager, watch out" 20:57 What Nurix is, and its Myntra origin 23:53 How big Nurix is today 24:47 Fermi: the AI tutor that won't give the answer 28:09 Back to Myntra: fashion then and now 30:06 How online became a liquidation channel 30:51 Koyu, Lyskraft and the CRED hypothesis 33:11 Why a venture studio, and why one company a year 39:09 Rejecting the compounding path 39:53 VC, valuation, and the near-death at Myntra 45:54 Haridwar, BHEL, and the books that lit the fire 49:06 Pivots as chess "adjustments" 51:39 Betting the megatrend, staying loose on the path 53:08 The Bay Area years and the return to India 57:37 Finding talent: "work with them for six months" 1:04:34 Learning vs winning: the live 2x2 1:07:51 Three books, and how he writes 1:14:23 Habits: sleep, the gym, one day a week to think 1:18:02 "No means not now" and other mental models 1:25:04 How health became a throughline 1:31:04 The longevity stack after 50 1:38:37 Why he runs SparX, and the deep-tech turn 1:44:05 Loyalty to a mission, and the number two 1:50:29 When AI supercharges the experienced 1:53:02 His kids' worldview, parents in Indian careers 1:56:55 Food, reading, and rating life an 8 Quotes [5:23] "I see myself as a learner. I do what I do because I enjoy learning… doing one company over a period of time, I get bored." [8:00] "Crisis always comes wrapped with a massive gift, if you are only willing to unpeel the onion and embrace the gift." [19:49] "For the first time, you can hire something equivalent to human cognition for cents an hour. Not using it is almost wasteful." [49:41] "I use the word adjustment rather than pivot. You come to work every day, you're playing chess." [1:18:xx] "No means not now. That's how I process all the nos in my life." (spot-check ts) Frameworks & mental models * Learner over operator: chase the zero-to-one learning curve; once a company becomes a treadmill of "how do I double this", hand it off. * Crisis and momentum: in a shock, cut the bad and double down on the good; "never let a crisis go to waste". * The three professionals in the AI age: the AI-native new-grad without judgment; the senior who has lost the ability to be hands-on; and the experienced operator who stays hands-on, for whom AI is a superpower. * Adjustment, not pivot: every day you make the best move on the board; most are small, occasionally you castle. A pivot that throws away the code, category and team is just a new company. * Bet the megatrend, stay loose on the path: be certain about the ten-year direction, zigzag the moves to get there. * Horizontal compounding: compound...
