@Sayak_Paul and I have worked on a project “automated TensorFlow model serving on Kubernetes with TensorFlow Serving and GitHub Action”, and the blog post is out on TensorFlow Blog today!
Here is the link
: LinkedIn
It assumes that the model is released as GitHub Releases in a designated GitHug repository, then GitHub Action detects the new release, then a new custom TensorFlow Serving docker image is going to be built, then it gets deployed on GKE(Google’s managed Kubernetes Engine).
Furthermore, you can learn how to manage and run multiple experiments with different sets of parameters for the TensorFlow Serving on kubernetes environment.
We hope you guys like it!
Thanks for reading 
- also special thanks to Robert Crowe and Josh Gordon for helping this to be pulished on TensorFlow Blog