Amazon Web Services on Tuesday unveiled Amazon Bio Discovery, a cloud-based artificial intelligence application designed to streamline early-stage drug development by allowing researchers to execute sophisticated computational workflows without writing code.
According to an AWS blog post, the platform gives scientists access to a curated library of specialized biological foundation models that can generate and assess candidate drug molecules. The service also incorporates an AI agent intended to help users choose the appropriate models, configure parameters and interpret the outputs produced by those models.
Workflows on the platform can progress from in silico design to experimental validation: researchers may submit shortlisted molecules to integrated laboratory partners for synthesis and testing, and experimental results are routed back into the system to inform further design iterations.
Rajiv Chopra, vice president of healthcare AI and life sciences at AWS, said the company believes the approach can markedly accelerate candidate generation. He noted that a process that previously required 18 months to produce roughly 300 potential drug candidates can now be completed within weeks. Chopra also pointed to a growing number of drug-discovery models and observed that this proliferation has created a shortage of computational biologists able to translate laboratory objectives into machine-learning pipelines.
AWS identified several early adopters of Amazon Bio Discovery, including Bayer (ETR:BAYN), the Broad Institute and Voyager Therapeutics (NASDAQ:VYGR). The company added that 19 of the top 20 global pharmaceutical companies are customers of its cloud services, underscoring existing relationships between AWS and major drug developers.
As part of its rollout, AWS will offer a free trial that includes five experimental units, with subscription pricing tiers to follow. In related activity at AWS's Life Science Symposium, the company said it will join Boston Consulting Group and Merck (NYSE:MRK) to present an AI platform aimed at improving clinical trial site selection.
The introduction of Amazon Bio Discovery highlights cloud providers' efforts to package machine-learning tools for scientific users who may lack deep computational expertise, while linking algorithmic design to laboratory execution and iterative feedback. The product launch is positioned to affect stakeholders across biotechnology, pharmaceuticals and cloud infrastructure services, given the combination of model access, orchestration and lab integration.
Note on limitations - The information above reflects AWS's description of Amazon Bio Discovery, its initial adopters and the trial offering as presented in the company's announcement.