Biomedical research is increasingly reliant on and driven by technologies that generate massive amounts of data to probe the mechanisms of disease and underpinnings of normal biology. The prime example of such a technology is high throughput sequencing applied to genomics research. In parallel with increased data generation capability, commercial cloud services have become commodity resources for flexible, scalable computing. Marrying cloud services, software tools for reproducible and portable workflows, and genomics data services and resources provides a powerful data ecosystem for large-scale genomic data science. Here, we present some key concepts, technologies, and software that allow for large-scale orchestration of data processing workflows and data science on commercial cloud resources. We also present some applications of these technologies to real-world genomics datasets, including how cloud computing can enhance portability, reproducibility, and data sharing. Finally, we highlight some of the challenges in adopting cloud technologies for biomedical data science.