Artificial intelligence continues to wave things in chemistry. Namely: Y combinator supported by Cambridge, based in the United Kingdom React Use AI to accelerate the manufacture of chemicals – a key step to put new drugs on the market.
Once a promising drug has been identified in the laboratory, pharmaceutical companies must be able to produce much higher quantities of equipment to conduct clinical trials. This is where Reactwise proposes to intervene with its “AI co -pilot for the optimization of chemical processes”, which says accelerates by 30x the standard process based on trials and error to determine the best method for making a drug.
“Make medication is really cooking,” said the co-founder and CEO Alexander Pomberger (photo above on the left, with the co-founder and CTO Daniel Wigh) in a call with Techcrunch. “You have to find the best recipe for making a medication with high purity and high performance.”
The industry has been based for years on what comes down to testing and error or error expertise for this “process development”, he said. The addition of automation in the mixture provides a means of reducing the number of iteration cycles necessary to land on a solid recipe to make a drug.
The startup thinks that it will be able to provide a “prediction of a single take” – where the AI will be able to “predict the ideal experience” almost immediately, without needing multiple iterations where the data on each experience are returned to refine the predictions – in the near future (“in two years”, is the bet of Pomberger).
Automatic learning AI models of the startup can always offer major savings by reducing the quantity of iteration necessary to exceed this piece of the drug development chain.
Boredom
“Inspiration was: I am a chemist in training, I worked in Big Pharma, and I saw how much the test and error are motivated throughout the industry,” he said, adding that the company essentially consolidates five years of academic research and doctorate focused on what it invoices as “simple software”.
Reactwise’s reactwise product is “thousands” of reactions that start -up has made in its laboratories to capture data points to feed its AI predictions. Pomberger says that the startup used a “broadband screening” method in its laboratory, which allowed it to detect 300 reactions at a time, allowing it to accelerate the process of capturing all these training data for its AI.
“In pharmacies … There are one or two handles of reactions, types of reactions, which are used again and again,” he said. “What we do is that we have a laboratory where we generate thousands of data points for these most relevant reactions, models of fundamental reactivity on our side, and these models can fundamentally understand chemistry.
The startup began this process of capturing the types of reaction to form its AIS last August, and Pomberger said that it would be completed by summer. He works to run 20,000 chemical data points to “cover the most important reactions”.
“To obtain a single point of data in a traditional way, it takes a chemist, generally, one to three days,” he said, adding: “It is therefore really, we call it, expensive to assess the data.
Until now, it focuses on the processes of manufacturing “small molecules drugs”, which, according to Pomberger, can be used in drugs targeting all kinds of diseases. But he suggested that technology could also be applied in other disciplines, noting that the company also works with two material manufacturers in the development of the delivery of polymer medicines.
Reactwise automation game also includes software that can interface with robotic laboratory equipment to further compose the manufacture of drug precision. However, to be clear, it is purely focused on the sale of software; He is not a manufacturer of robotic laboratory kit himself. On the contrary, he adds another string to his bow to be able to offer to drive robotic laboratory equipment if his clients have such a kit to put.
The British startup, founded in July 2024, has 12 pilot trials of its software in progress with pharmaceutical companies. Pomberger said they expected the first conversions – in large -scale deployment of the subscription software – later this year. And although he does not yet reveal the names of all the companies with which he works, Reactwise says that these trials include certain major pharmaceutical players.
Pre-series financing
Reactwise reveals all the details of its pre-series increase, which totals $ 3.4 million, said the startup exclusively.
The figure includes the previously disclosed support of YC ($ 500,000) and a Innovate UK Grant of almost 1.2 million pounds sterling (approximately $ 1.6 million). The rest of the funding (around 1.5 million dollars) comes from unnamed venture capital and providential investors, which, according to Reactwise, are “determined to advance sustainable and sustainable pharmaceutical manufacturing”.
While Reactwise is concentrated, quite closely, on a specific part of the drug development chain, Pomberger said that acceleration here can make a significant difference by shrinking the time necessary to obtain new pharmaceuticals to patients.
“Let’s take a typical duration of a drug from start to launch: 10 to 12 years.
Simultaneously, other startups apply AI to different aspects of drug development, including the identification of interesting chemicals in the first place, so there are probably composition effects as more and more automation innovations are folded down.
But with regard to the manufacture of drugs, in particular, Pomberger maintains that Reactwise is ahead of the pack. “We were the first to approach this,” he said.
The startup competes with the software inherited using statistical approaches, such as JMP. He also said that there were a few other people applying AI to accelerate the manufacture of drugs, but said that Reactwise’s access to high -quality data sets on chemical reactions gives it the competitive advantage.
“We are the only ones to have the capacity of, and which currently generate these high quality data sets internally,” he said. “Most of our competitors, they provide the software.
“But, on our side, we offer these pre-trained models-and these are extremely powerful because they fundamentally understand chemistry.
