A San Francisco SaaS company quietly rewiring how the world's specialty chemicals get invented, tested, and shipped.
Photographed in pixels, somewhere between a wet lab in Charlotte and a server rack in us-west-2.
Walk into a specialty chemicals lab in 2026 - one of the good ones, the kind that supplies the coating on your phone or the binder in a cement floor - and you will probably still see a paper notebook. Or worse: a chemist alt-tabbing between Excel, a vendor portal, and a 14-year-old LIMS interface that was last loved in the second Obama administration. This is the industry Alchemy Cloud has decided to fix.
The company calls itself an "AI-native Applied Sciences Platform." That is the polite version. The honest version is that Alchemy is trying to give every formulator on Earth one place to think - a single cloud system where the experiment, the sample, the analytical result, the customer request, and the predicted property all live next to each other. ELN, LIMS, PLM, and AI in one product, instead of four vendors and a Friday-afternoon prayer.
Today the platform sits inside roughly fifty industrial heavyweights - Milliken, RPM, Carboline, Tremco, AGC, Arxada, IMCD - companies whose names you have never tweeted about but whose products are quietly in your kitchen, your car, your office floor, and the glass on your office window. About 53 people work on it from San Francisco, New York, and a quiet engineering office in South-Central Europe.
Specialty chemicals is a $1 trillion global business. It is also, by most honest measures, one of the least digitized corners of the industrial economy. Formulations get tracked in spreadsheets. Test results get emailed. Decades of R&D sit on hard drives belonging to chemists who have, politely, retired.
The cost shows up everywhere. New products take years instead of months. Customer technical-service teams cannot find data their own lab generated last quarter. The same experiment gets run three times in three sites because nobody knew. And good luck training an AI model on a dataset that lives in eighteen file formats.
The other complication is that fixing this is genuinely hard. A lab is not a CRM. Experiments have units, instruments, hazards, raw-material variances, and regulatory implications. A "minor refactor" can render a product non-compliant in three countries. This is why the easy software solutions - the ones that look great in a sales demo - usually die six months after rollout.
Alchemy was founded in 2017 by Sasha Novakovich, Dusko Vesin, and Nikola Milinkovic. The unusual move was the org chart. Most startups have one CTO and a lot of optimism. Alchemy started with two - one CTO running ELN and LIMS, another running AI and DOE - because the founders did not believe one head could honestly hold both halves of the problem.
It was, on paper, a slightly ridiculous bet. Selling cloud software to a conservative industry that had survived a century without it. Replacing entrenched legacy systems with multi-million-dollar implementation budgets behind them. Convincing a regulated lab director to put their crown-jewel formulation data on someone else's computer.
The commercial brain. Sells to scientists in their own language.
Owns the digital lab bench. Makes the software a chemist will actually open on Monday.
Runs the model layer. Turns lab data into predictions chemists can argue with.
Investors agreed enough to put $9.4 million in over multiple rounds, the most recent a $3M cheque in August 2022. Not a hype number. The kind of number you raise when you intend to run a long, capital-efficient game inside an industry that will not forgive a startup for blowing up its data.
It is fashionable to say "we are a platform." It is also usually a lie. Alchemy gets a pass because it actually has to be one - chemistry data refuses to live in silos.
Cloud-native notebook for chemists, with chemical drawing, structured experiments, and real-time collaboration. Replaces paper and spreadsheets without asking anyone to learn Python.
Sample tracking, instrument integration, analytical test workflows - all wired to the same data model the chemist already uses.
Walks a formulation from first experiment to commercial product to customer-facing technical data sheet. One spine, not five.
Property prediction, formulation recommendations, automated Design of Experiments. The kind of AI that makes a senior chemist mutter "huh" rather than "no."
Links technical-service tickets and customer samples back to the R&D record - so the sales team can finally answer a customer in hours instead of weeks.
Founded in San Francisco by three co-founders.
Early specialty-chemicals design partners; ELN and LIMS shipped.
PLM module released; enterprise customer base crosses a dozen.
$3M venture round closed; total raised reaches $9.4M.
AI for property prediction and DOE automation expands.
50+ industrial customers; offices on three continents.
The cleanest test for an enterprise SaaS company is not how loud it is on LinkedIn. It is which logos quietly renew. Alchemy's list reads like an industrial periodic table.
Directional estimate, based on Alchemy's publicly cited verticals.
Customers include Milliken, RPM, Carboline, Tremco, AGC, Arxada, and IMCD - the kind of companies whose chemistry quietly holds up large parts of the physical world.
Alchemy also keeps a small set of partnerships where they need to - including interoperability with LabWare LIMS environments, because a healthy chunk of the industry already runs on it and a rip-and-replace conversation rarely ends in a signed contract.
That is the literal mission statement on their website. It is the kind of line that would sound corny coming from a Series F unicorn. Coming from a 53-person team that sells to coatings companies, it lands differently - because the chemistry under that sentence is the chemistry that makes batteries last longer, food safer, and buildings less flammable.
The mission also explains the product strategy. You cannot accelerate science by selling a prettier notebook. You accelerate it by making every experiment in the world connected, searchable, and - increasingly - predictable. That is what an AI-native lab actually means. Not "we added a chatbot." It means: when a chemist starts a new project, the platform already knows what worked, what failed, and what is worth trying next.
Hardware ate software, then software ate hardware, and the next wave is software eating the wet lab. Battery breakthroughs, sustainable plastics, food-grade preservatives, lower-carbon cement - none of these get faster without a digital backbone underneath them.
Alchemy's bet is that the company holding that backbone, for the parts of the economy that actually make physical things, is going to be worth a lot. Not because of hype. Because the underlying problem is large, the customers are sticky, and there is no clean second-place vendor that does ELN, LIMS, PLM, and AI under one roof.
There are risks, of course. Enterprise sales cycles are long. AI in regulated industries is sensitive. The next downturn will test which customers genuinely treat lab software as critical infrastructure. But the direction of travel is hard to argue with: the chemist of 2030 is not going back to paper.
So go back to the opening scene. The specialty chemicals lab. The paper notebook. The chemist alt-tabbing between four windows. That is the image Alchemy Cloud is in the business of replacing - one customer, one formulation, one query at a time. The notebook closes. The browser tab opens. The data starts talking back. Quietly, and on schedule, an industry that powered the twentieth century is being rewired for the twenty-first.