Biotech firm XtalPi Holdings, which specializes in accelerating drug development, has increased its success rate for chemical experiments from 20–30% to 90% by using a specially designed AI model, SCMP reports.

“There are many good opportunities [for AI] in vertical fields. In the pharmaceutical industry, we see great potential for LLM applications,” said Zhang Peiyu, XtalPi’s chief scientific officer.

He predicts that integrating robotics and artificial intelligence could reduce the drug discovery timeline from four years to two.

XtalPi was founded in 2014 by quantum physicists from the Massachusetts Institute of Technology. A year later, the company set up its research and development base in Shenzhen, China.

Today, XtalPi is a key player in AI-driven drug development, serving nearly 80% of the world’s major pharmaceutical firms. Its automated chemistry laboratory houses approximately 200 AI-powered robots.

Despite the firm’s notable achievements, Peiyu pointed out that China still lags behind the United States in top-tier innovation and fundamental research in biomedicine.

Context

In December, researchers at the Chinese University of Hong Kong developed an AI model for clinical ophthalmology tasks such as screening and diagnosing diseases.
Earlier, scientists created a “smart throat” system that helps stroke and dysarthria patients regain natural speech.