Speaker
Description
Science workloads increasingly rely on cloud platforms to provide the elastic compute, storage, and networking needed for data- and compute-intensive research. By leveraging cloud-native technologies such as containers, Kubernetes, and managed services, scientific applications can scale from interactive prototyping to large parallel experiments while improving reproducibility and portability across institutions. This shift enables researchers to co-locate data and compute, automate complex workflows, and rapidly adopt accelerators and specialized hardware without owning physical infrastructure. At the same time, it introduces new challenges in performance variability, cost optimization, security, and compliance with data governance policies.
One key and often overlooked aspect of working in the cloud is staying up to date with all the various tools and technologies. Kubernetes, which is the de facto orchestration platform for running containerized applications at scale, follows a four-month release cycle, where each release reaches end of life after 14 months. Organizations benefit from clear version policies that define supported releases, planned upgrade windows, and systematic testing of breaking changes in staging environments. Continuous learning—through hands-on labs, workshops, certifications, and community engagement—helps teams safely adopt new capabilities reliably.