AI is making waves once again in the world of chemistry and drug manufacturing. Cambridge-based startup ReactWise, a Y Combinator-backed company, is harnessing artificial intelligence to dramatically accelerate chemical manufacturing—a crucial stage in bringing new drugs to market.
Once a drug candidate shows promise in the lab, pharmaceutical companies face the challenge of scaling up production for clinical trials. This step, known as process development, often slows down drug timelines due to its heavy reliance on trial-and-error methods. ReactWise aims to change that with its “AI copilot for chemical process optimization.” The platform claims to speed up this phase by a staggering 30 times.
“Making drugs is a lot like cooking,” explained ReactWise co-founder and CEO Alexander Pomberger. “You need to find the best recipe that yields the highest purity and output.” Traditionally, chemists depend on experience and repetitive testing to refine these “recipes.” ReactWise introduces automation and machine learning models that reduce the number of iterations needed, cutting down both time and costs.
The startup is pushing toward what it calls “one-shot prediction” — an ambitious goal where AI can instantly suggest the ideal chemical experiment without multiple rounds of testing. Pomberger is optimistic this could be achievable within two years.
ReactWise’s AI models are built on thousands of in-house chemical reactions. By running high-throughput screenings of up to 300 reactions at a time, the team generated rich datasets to train their AI. Their focus is on small molecule drugs, commonly used in treatments across various diseases. Pomberger highlighted that a single chemical data point traditionally takes one to three days to produce, making their automated system a major breakthrough.
Beyond pharmaceuticals, ReactWise sees potential applications in materials science, such as polymer drug delivery. Their software can also integrate with robotic lab equipment, giving clients the flexibility to automate further if they have the infrastructure. However, ReactWise remains focused on software, not hardware manufacturing.
Founded in July 2024, the U.K. startup has already launched 12 pilot trials with pharmaceutical companies, including undisclosed Big Pharma players. Full-scale subscriptions are expected later this year as trials progress.
ReactWise also announced its pre-seed funding round totaling $3.4 million. The funding includes $500,000 from Y Combinator, nearly £1.2 million from Innovate U.K., and around $1.5 million from VCs and angel investors passionate about AI-driven, sustainable pharma manufacturing.
Pomberger believes that ReactWise’s approach could significantly shorten drug development timelines. “Process development takes one to two years in a typical 10 to 12-year drug launch cycle. If we can reduce that by 60%, the impact is huge,” he said.
While many AI startups focus on identifying promising drug manufacturing compounds, ReactWise stands out by targeting the drug manufacturing process. Competing mainly against legacy statistical software like JMP, ReactWise claims its proprietary, high-quality datasets give it a distinct advantage.
“We offer pretrained models that truly understand chemistry,” Pomberger added. “Our clients can input their reaction of interest, hit start, and receive process recommendations immediately. No one else provides this level of readiness right now.”
