From statistics to agentic AI, in seven chapters
Fourteen years across five industries. The convictions I write about came from here.
- Chapter 12007–2012
Foundations
BHU · IIT Bombay · B.Sc. Statistics · M.Sc. Applied StatisticsI played cricket at BHU and did student politics at IIT Bombay, and was just intelligent enough to score 7 out of 10 at both.
- Chapter 22012–2014
Models that make sense
WNS · Senior AnalystMy first job taught me that a model earns trust only when the business can see itself in it. Demand never moved on price alone. It moved with how people felt, and a forecast that ignored that was just arithmetic.
- Chapter 32014–2017
Measuring the new
Citi · Assistant ManagerA bank taught me that what customers say and what they do are connected in ways no single metric shows. And when a channel is new, whoever measures it honestly first shapes how everyone spends on it.
- Chapter 42017–2018
A short stop
Optum · UHG · Manager, Data SciencesA short stop, but long enough to see that when a model's output touches someone's health cover, roughly right is not good enough.
- Chapter 52018–2020
The turning point
Walmart Labs · Data ScientistThis is where I turned into a hardcore applied scientist. Retail at that scale is unforgiving: one decision in one store ripples through demand, inventory and suppliers, and the rigour has to match it.
- Chapter 62020–2025
The awakening
Koch · Data Scientist → Data Science LeadTaking AI from slides into production woke me up to one thing: the model is the easy part. The engineering around it, the connections between decisions, and whether people actually use it decide everything.
- Chapter 72025–now
Owning the whole cycle
Aditya Birla Group · Senior Analytics Practice LeaderNow I own the whole cycle, from the first strategy conversation to adoption on the ground. It is where I saw most clearly that AI pays only when a business is willing to redesign how it works around it.