For Daniil Boiko and Andrei Tyrin, the idea for Onepot AI it came from the same frustration.
“The best ideas for drug discovery were often held back not by biology, but by synthesis,” Boiko told TechCrunch. Synthesis is the creation of new molecules using chemical reactions. It is like a recipe or Lego pieces, where small pieces, ingredients, molecules, come together to form a larger puzzle picture, a plate of food, a larger molecule.
As you might expect, it is quite difficult to create those small molecules that go on to create larger ones.
For Boiko, a doctoral candidate studying machine learning in chemistry at Carnegie Mellon (he got his bachelor’s and master’s degrees in organic chemistry from a university in Russia), this meant that drug hunters — the scientists who oversee drug discovery and development — were passing up promising ideas just because the chemical molecules seemed too difficult to create the drug.
“The compounds have never had a chance to be tested,” Boiko told TechCrunch.
For Tyrin (who received his bachelor’s degree in computer science at MIT), his time working on computational drug discovery pipelines made him realize how far behind the world of drug discovery was. “The models could generate ideas in hours, but it could take months to reach the lab,” he told TechCrunch.
“We both saw that people were throwing money at molecular design and almost ignoring the harder problem of actually making the molecules,” Boiko said. But there was also a geopolitical angle, he continued, with global supply chains becoming vulnerable and the US entering a trade war and innovation competition, once again, with China.
Techcrunch event
San Francisco
|
13-15 October 2026
“It was clear,” Boyko said. “Small molecule synthesis had to be rebuilt from the ground up in the United States.”
Boiko and Tyrin teamed up to create Onepot, a company that hosts the POT-1 small molecule synthesis lab. They also built Phil, an AI organic chemist, to help conduct experimental analysis to scale up the compound synthesis process for their first commercial partners. These partners are biotech and pharmaceutical companies that are currently testing their technology.
On Wednesday, the company emerged from secrecy with $13 million in funding, including pre-seed money, and a seed round led by Fifty Years.
“Currently, pharmaceutical and biotech companies are either building entire teams of chemists in-house or working with contract research organizations overseas,” Tyrin said of the process for molecular synthesis. Human chemists can spend months of research to create even one compound, at a cost of thousands of dollars.
It involves a lot of trial and error — studying various compounds, gathering data on biological activity, how the drug moves through the body, toxicology reports, and deciding what to experiment with next. “The main limiting factor here is not testing these compounds, but making them in the first place,” Tyrin continued. “We plan to squeeze it into days.”
Tyrin said the product is pretty simple. Onepot has a list of molecules it can make. Customers choose which compounds they want, and then Onepot’s technology will synthesize the molecules and ship them to the customer for the customer to use in their own experiments. (They ship natural products either as dry compounds or as solutions in plates or vials.)
The back of the product is where Boiko and Tyrin have fun, analyzing the problems of chemical synthesis to discover which combinations of molecules work together. They built a lab where they let LLM agents have access to these so-called molecule recipes for training, so agents can also discover what works and what doesn’t in complex building.
“When performing experiments in the lab, we record every detail involved in the process,” Tyrin said—that means tracking the temperature and, essentially, the ingredients added to a mixture to create compounds. “No information is lost, which makes the experiences reproducible even if someone decides to run them 10 years from now.”
This also means that their agents generate hypotheses from real-world experiments rather than literature data, often mined from the Internet.
Boiko called the fundraising process “tumultuous” and said they met their lead investor through an introduction. “What was supposed to be a short meeting turned into an hour-long whiteboard session on the industrialization of synthesis,” Boyko said. Others in the round include Khosla Ventures, Speedinvest, OpenAI co-founder Wojciech Zaremba, and Google chief scientist Jeff Dean.
The fresh capital will be used to build a second lab in San Francisco so the team can take on more clients. It will also expand its compound discovery team and engine. On the services side, Boiko and Tyrin see WuXi AppTec and Enamine as competitors.
Overall, Boiko and Tyrin hope to make drug discovery at least twice as fast, and hopefully change the perception of what’s possible once someone gets hold of the “strange” chemistry that scientists once thought was off limits.
“You’re not just accelerating drug discovery, you’re expanding the design space for what drugs and materials can be,” Boiko said. “That drug that we haven’t discovered yet might be out there waiting for us to find it.”
