With the recent advancements in AI/ML and High Performance Computing, businesses have started building compute-intensive applications that require distributed and parallel execution. Ray is an open source project/framework (developed at UC Berkeley RISE Lab) that enables distributed execution of compute-intensive Python workloads. The talk focuses on architecting and running large-scale compute-intensive distributed Python applications with Ray on Kubernetes by leveraging Kubernetes orchestration, auto-healing, auto-scaling, security, and observability.