Mstack

Shreyans Chopra spent three years at Lightspeed India writing checks into other people's companies, after a stint at Boston Consulting Group and a Chartered Accountant qualification. This is a fairly standard resume for someone who ends up starting a fintech app. It is a strange resume for someone who ends up running wet labs. In 2022 Chopra left investing to found Mstack, a company whose product is, at the most literal level, molecules in barrels.

The problem Mstack set out to solve is one that got a lot louder in the 2020s. Western manufacturers - the ones making agrochemicals, coatings, water-treatment additives, oilfield chemicals - increasingly wanted to diversify away from Chinese suppliers. European producers were squeezed by energy costs. The natural alternative was a long tail of capable manufacturers in India and the Middle East. The catch: a procurement manager in Houston has no easy way to know whether a mid-sized plant in Gujarat will deliver the right purity, on time, at the quoted price. Trust does not travel well across borders.

Mstack's answer is to be the trusted party in the middle. Not a passive marketplace where buyers and sellers find each other and hope for the best, but an active partner that sources, tests, ships, delivers, and tracks - and puts its own name on the outcome. The pitch to the enterprise buyer is refreshingly unglamorous: you get one platform and one throat to choke, instead of a dozen vendor relationships and a lot of faith.

It worked quickly. By the time Mstack raised money in late 2024, Chopra was telling reporters the company had grown sales roughly fifteenfold in a year and served more than fifty enterprise clients across six geographies. Which brings us to the part where a former Lightspeed investor took money from Lightspeed.

In October 2024 Mstack announced a $40 million Series A, co-led by Lightspeed and Alpha Wave Global, with a debt facility from HSBC Innovation Banking and a handful of angels. Lightspeed partner Bejul Somaia talked about transforming the specialty-chemicals market; Alpha Wave co-founder Navroz Udwadia talked about reshaping global chemical supply chains. This is the sort of language investors use when they have decided a large, unsexy market is about to get software.

A small accounting note, because it is the kind of thing worth being honest about: some datasets record the round as roughly $32 million, while the announced headline figure was $40 million. The likely reconciliation is that the larger number blends equity with the HSBC debt facility. Either way, the money was earmarked for expansion across North America, Latin America, the Middle East, and Asia, plus building out R&D capacity in India - including a Hyderabad lab now staffed by more than 25 scientists.

The genuinely unusual financial detail arrived a year later. When Mstack launched its AI platform in September 2025, it reported not just 10x revenue growth but operational profitability. In an era where "AI company" is often shorthand for "burns cash impressively," a business that makes physical goods and turns an operating profit while growing fast is the contrarian trade. The atoms, it turns out, pay for the algorithms.

Most companies that call themselves "AI-native" would be delighted never to touch the physical world. Mstack does the opposite, on purpose. It runs a Hyderabad wet lab. It answers to Baker Hughes and Veolia. It ships product that either reproduces in a customer's plant or does not. This is harder. It is also, arguably, the whole point - because a synthesis model that has never been checked against a real reactor is just an opinion with good grammar.

There is a nice symmetry to the founding story, too. Chopra was an investor who kept looking for a company like this and, not finding one, built it - then took money from the firm where he used to work. The team he assembled is deliberately mismatched: machine-learning engineers and bench chemists and supply-chain operators, spread from San Francisco to Bangalore to Hyderabad, arguing in the same rooms. The thesis is that neither discipline wins the chemistry problem alone.

Whether the timeline claims hold up outside the demo is the open question, and worth watching rather than swallowing. "Eighteen months to days" is a headline; the honest version is "for the molecules and routes where our models are strong, dramatically faster." But even the conservative reading is a real change to how specialty-chemical R&D gets done. And the financials - fast growth, operational profit, physical goods - are unusual enough to make Mstack one of the more grounded stories in a field full of promises.