The accountant who cold-emailed his way in - and now wants AI to be the world's smartest chemist
Shreyans Chopra couldn't crack the rank list. He talked his way into BCG anyway, left a venture capital seat, and built Mstack - an AI-native chemicals company that hit profitability in a market most founders won't touch.
There is a moment in most careers where the rulebook says stop. For Shreyans Chopra it arrived when he finished his chartered accountancy exams in Ahmedabad without a rank. In India's consulting pipeline, the CA final rank list is a velvet rope. No rank, no interview. Most people accept that. Chopra wrote an email instead. The subject line, sent to the India heads of BCG and McKinsey, read: "I am not a CA final ranker." BCG called him in.
That email is the whole man in miniature - direct, a little cheeky, allergic to the idea that a gate is really closed. Today he runs Mstack, a specialty chemicals company headquartered in Houston that describes itself as AI-native. It raised a $40 million Series A in 2024, reports 10x revenue growth, and - unusually for anything with "AI" in the pitch - says it is operationally profitable. He was 29 when Forbes India put him on its 30 Under 30 list.
A Marwadi kid who took the long way around
Chopra grew up in Ahmedabad in a Marwadi business family - the kind of household where commerce is discussed at the dinner table rather than taught in a classroom. His father traded aluminium. Extended family ran small manufacturing outfits. His two older siblings both went to IIT, the default prestige path. Chopra took the accountancy route instead, following a family tradition, interning at Deloitte and grinding through the CA program at G.K. Choksi & Co.
Numbers were the day job. Music was the pull. He had been captain of his college band, and after graduating he tried to turn that into a business called Funkaar Gigs - a platform connecting more than 300 Ahmedabad musicians with venues and event organizers. It worked, sort of. It also didn't scale. He shut it down after concluding, with characteristic bluntness, that it "looked like a college project." That instinct - the willingness to kill something he loved because the math didn't work - matters more than the failure itself.
- 2013Interns at Deloitte; becomes a chartered accountant at G.K. Choksi & Co.
- 2016Co-founds Funkaar Gigs, booking live musicians in Ahmedabad
- 2017Cold-emails his way into Boston Consulting Group
- 2019Joins Lightspeed India as an early-stage investor
- 2022Founds Mstack in July
- 2024Forbes India 30 Under 30; $40M Series A
- 2025Launches Chemstack AI; reports profitability
Accountant, band captain, consultant, investor, founder. Not a straight line - which is rather the point.
The investor who wanted to build
BCG led to two years of consulting, which led to Lightspeed India, where Chopra spent three years as an early-stage investor. It is a coveted seat. Most people who reach it stay. Chopra used it as a listening post. Sitting across the table from founders every week, he formed a conviction that the most interesting opportunities were not in another consumer app but in the enormous, unglamorous, offline industries that software had barely touched.
Specialty chemicals is about as unglamorous as it gets - and about as large. Grand View Research pegs the global market near $650 billion. It is fragmented, opaque, and heavily concentrated in a handful of producing countries. When Chopra looked at it, he saw a market desperate for exactly the thing his whole career had prepared him to build: trust across borders.
He left Lightspeed and founded Mstack in July 2022. The plot twist that even a screenwriter would flag as too neat: two years later, Lightspeed co-led his Series A. The seat he gave up became the check he received.
Not a marketplace, on purpose
The easy version of Mstack would have been a listings site - a place where buyers and chemical suppliers find each other and the platform skims a fee. Chopra rejected it. A thin brokerage layer, he argues, cannot promise the one thing enterprise buyers actually need: that the drum arriving in Houston contains exactly what the spec sheet says. So Mstack sits inside the supply chain. It sources from a tech-enabled network of manufacturers across Asia and the Middle East, controls testing and quality, and handles shipping, delivery and tracking - a single trusted brand rather than a bulletin board.
The timing helped. As tariffs of 10 to 50 percent landed on chemical intermediates entering the United States, and as Western buyers grew nervous about leaning so heavily on China, Mstack's pitch - a diversified network spread across six geographies - stopped being a nice-to-have. Chopra reads the macro anxiety and sells the calm on the other side of it.
Sales grew roughly 15x in the twelve months to October 2024; by 2025 the company reported 10x growth year-on-year and operational profitability. Bars are illustrative of the trajectory, not audited figures.
Teaching software to run the lab
The commerce engine is one half of Mstack. The other, launched in September 2025, is Chemstack AI - an attempt to compress the slowest part of chemistry. Developing and commercializing a new molecule traditionally takes something like 18 months. Mstack claims it can get there in days. The system is a closed loop of three parts, each doing a job a human team would otherwise slog through by hand.
Read the science, design the route, run the experiment - then feed the results back in. The lab that never stops learning.
Behind the software sit real benches: R&D laboratories in Hyderabad staffed by 25 or more scientists, feeding data back into the models. The company now serves customers across North America, India, China and the Middle East, spanning CASE coatings, agrochemicals, oil and gas, personal care and construction chemicals, with a team of more than 100 across data science, R&D, operations and sales.
The pattern, if you want to steal it
There is a repeatable move underneath Chopra's story, and it is not about chemicals. It is about refusing the framing you're handed. The rank list said no; he changed the question. The marketplace was the obvious model; he built the harder, thicker one. AI startups are supposed to burn cash chasing scale; he chose to make money selling something you can hold in a drum. Each time, the crowd optimized for the easy version and he took the version with a moat.
He is careful about the numbers - revenue figures stay close to the chest - and he talks like someone who has watched enough pitches from the other side of the table to know the difference between a narrative and a business. Whether Mstack becomes the industry-defining company Forbes India suggested it might is unsettled. But the bet is clear, and it is his: that the least fashionable corner of the physical economy is exactly where an AI-native company should plant its flag.
For a kid whose siblings took the IIT escalator and whose first company booked bands, that is a long way to travel on the strength of a cold email. He would probably tell you the email was the easy part. The hard part was believing the door would open.