Run AI at scale with Ray + Kubernetes using Anyscale on Azure | ODSP914
Anyscale on Azure brings Ray-powered distributed AI compute to AKS for owning your AI stack at scale
“It has over 12 million downloads per week.”
At Microsoft Build, Anyscale introduced Anyscale on Azure, powered by Ray on AKS, offering a unified runtime for multimodal data curation, distributed training, and multi-model serving for teams that want to own rather than rent their AI stack. The pitch centers on Ray's growing dominance (12M+ weekly downloads, used by OpenAI and Uber) as the convergence point for scaled AI workloads. It matters as a vendor signal that enterprises are increasingly moving toward owning open-source-based AI infrastructure, but it is largely a marketing/product session rather than a major industry shift.