Ashok Chandrasekar

Tech lead for performance evaluation and optimization of AI inference workloads on GKE. - Built a benchmarking tool for inference performance evaluation on Kubernetes - https://github.com/GoogleCloudPlatform/ai-on-gke/tree/main/benchmarks - Launched the Gemma model on GKE - https://cloud.google.com/kubernetes-engine/docs/tutorials/serve-gemma-gpu-vllm - Improved model serving performance via accelerator / model server optimization and through intelligent autoscaling Previously, served as the Tech Lead for the Node Management area where I helped improve disruption management of GPUs and TPUs and made node pool updates and upgrades simpler and reliable.

Tech Lead for the Cluster Lifecycle Management team in Tanzu Mission Control - https://tanzu.vmware.com/mission-control

Worked on VMware Cloud PKS: - Developed the Kubernetes cluster lifecycle management component - Worked on improving the security of Kubernetes clusters - Worked on automatically scaling the Kubernetes clusters based on workload - Worked on managing storage in Kubernetes clusters

Designed and developed a distributed scheduler for Photon Controller which used shared-state optimistic concurrency control to improve performance with the help of Xenon (https://github.com/vmware/xenon).

Worked on a highly scalable IaaS cloud platform Photon Controller (https://vmware.github.io/photon-controller/) where I worked on developing the deployer which used a single node control plane to bootstrap itself by deploying a multi-node highly available control plane.

Teaching assistant for the Advanced Cloud Computing course where I was in charge of creating the project that students worked on and giving a lecture occasionally. Project that I developed was to perform page ranking using Hadoop MapReduce and Apache Spark and to compare the effectiveness of using both the systems for batch processing vs real time workloads.

Built a datacenter resources importer which can import resources from multiple clouds (OpenStack, vSphere) for a highly scalable IaaS cloud platform.

Teaching assistant for the Distributed Systems course where I helped develop and mentor students on their Distributed Systems projects.

Developed a talent management web application using Ruby on Rails for enterprises with two other engineers.