For Daniil Boiko and Andrei Tyrin, Onepot AI was born from a shared frustration.
“Often, biology wasn’t the hurdle for the best drug discovery ideas; synthesis was,” Boiko explained to TechCrunch. Synthesis involves the creation of novel molecules via chemical reactions. Think of it as a recipe or Lego set: smaller components (ingredients, molecules) unite to form something bigger (a food dish, a larger molecule).
Unsurprisingly, producing the smaller molecules required to assemble larger ones is challenging.
For Boiko, a Ph.D. candidate in machine learning for chemistry at Carnegie Mellon (with bachelor’s and master’s degrees in organic chemistry from a Russian university), this meant observing that drug hunters—scientists leading drug discovery and development—were passing over promising concepts simply because the necessary chemical molecules seemed too difficult to produce.
Boiko told TechCrunch, “These compounds never even had the opportunity to be tested.”
Tyrin (who holds a bachelor’s in computer science from MIT) realized, through his work on computational pipelines for drug discovery, how far behind drug discovery was. He told TechCrunch, “The models could generate ideas in hours, but it could take months for the lab to keep pace.”
“We both noticed that investments were heavily skewed towards molecular design, with the more challenging issue of molecule creation largely overlooked,” Boiko stated. He added that there was also a geopolitical aspect: Global supply chains are becoming more at risk, and the U.S. finds itself in another trade war and innovation race with China.
Techcrunch event
San Francisco
|
October 13-15, 2026
Boiko stated, “It was obvious: Small molecule synthesis needed a complete overhaul within the United States.”
Boiko and Tyrin collaborated to form Onepot, which houses the small-molecule synthesis lab POT-1. They also developed Phil, an AI organic chemist who aids in experimental analysis to enhance the compound synthesis process for their initial commercial partners. These partners include biotech and pharma companies testing their technology.
On Wednesday, the company emerged from stealth with $13 million in funding, including pre-seed funds, and a seed round spearheaded by Fifty Years.
Tryin commented on the molecular synthesis process, saying, “Currently, pharmaceutical and biotech firms either assemble full teams of in-house chemists or collaborate with overseas contract research organizations.” Human chemists might spend months researching to produce just one compound, at a cost of thousands of dollars.
This involves considerable trial and error — examining various compounds, gathering data on biological activity, how the drug behaves in the body, toxicology results, and determining what to experiment with next. Tyrin continued, “The primary constraint isn’t testing these compounds but creating them in the first place. We want to shorten this to days.”
Tyrin described the product as quite simple. Onepot offers a catalogue of synthesizable molecules. Clients select their desired compounds, and then Onepot’s technology synthesizes the molecules, shipping them to the customer for use in their own experiments. (Physical products are shipped as dry compounds or solutions in plates or vials).
Boiko and Tyrin enjoy delving into the chemical synthesis challenges, figuring out which molecular combinations are effective. They’ve established a lab where LLM agents can access these molecule recipes for training, allowing them to determine what succeeds and fails in compound construction.
Tyrin explained, “We meticulously record every detail of lab experiments, ” including temperature monitoring and tracking the ingredients added to create compounds. “No information is lost, ensuring experiments are reproducible even after a decade.”
This also means that their agents formulate hypotheses based on actual experiments rather than internet-sourced literature data.
Boiko described the fundraising as “hectic,” adding that they met their lead investor through an introduction. Boiko recalled, “What began as a brief meeting evolved into an extensive whiteboard session on industrializing synthesis.” The round also included Khosla Ventures, Speedinvest, OpenAI co-founder Wojciech Zaremba, and Google’s Chief Scientist Jeff Dean.
The new funds will support the construction of a second San Francisco lab to accommodate more clients. The team and its compound discovery engine will also expand. Regarding service competition, Boiko and Tyrin consider WuXi AppTec and Enamine as rivals.
Overall, Boiko and Tyrin aim to accelerate drug discovery by at least twofold and transform perceptions of what’s achievable by leveraging the “unconventional” chemistry previously deemed off-limits.
Boiko asserted, “It’s not just about accelerating drug discovery; it’s about broadening the design possibilities for drugs and materials. The drug we haven’t yet discovered could be out there, waiting.”
