Hey Manoj, I am not sure how clearml-session would know how to setup kube-proxy, if that's your intent.

Personally, I would run the clearml-server and agents on the k8s cluster, and then expose the server endpoints via kube proxy or some other nicer ingress. Then you can run jupyter locally and you should be good. Jupyter session remotely running on k8s would be a logistical nightmare 🙂