Amazon Launches Bio Discovery Platform to Streamline AI-Driven Drug Research

The platform introduces a “lab-in-the-loop” approach, where AI-driven predictions are continuously refined using experimental feedback from laboratory testing.

Amazon has launched Amazon Bio Discovery, a new application aimed at addressing workflow inefficiencies in drug discovery by combining computational design and wet-lab validation into a unified system.

The platform introduces a “lab-in-the-loop” approach, where AI-driven predictions are continuously refined using experimental feedback from laboratory testing. This model is intended to reduce delays caused by disconnected workflows between computational biologists and bench scientists.

Amazon Bio Discovery provides access to more than 40 AI biology models, allowing researchers to select, evaluate, and deploy models based on specific research needs. The system includes agent-based assistants that guide users in choosing models, configuring experiments, and analyzing results without requiring coding expertise.

The application also allows researchers to upload custom or third-party licensed models and build reusable workflows within a no-code environment. These workflows standardize data processing and analysis, enabling teams to scale research programs without additional infrastructure management.

A key feature of the platform is its integration with contract research organization (CRO) partners, including Ginkgo Bioworks, Twist Bioscience, and A-Alpha Bio. Researchers can directly send selected drug candidates for wet-lab validation, with cost estimates and turnaround times generated within the system. Experimental results are automatically fed back into the platform to improve subsequent modeling cycles.

Amazon said the platform has already been used in collaboration with Memorial Sloan Kettering Cancer Center to design nearly 300,000 antibody candidates, with top candidates shortlisted and sent for testing within weeks, compared to traditional timelines that can extend up to a year.

The system also includes tools for molecular analysis, candidate filtering, and validation checks, enabling researchers to assess drug properties such as stability, efficacy, and developability before lab testing.

Amazon Bio Discovery is built on AWS infrastructure and includes data isolation features to secure proprietary research data and models. The platform is now available, with access options including a free trial and training resources through AWS.

Stay tuned for more such updates on Digital Health News