Becoming cloud-agnostic through Kubernetes
One of the key aspects of the ML platform we are building is that it enables the organization to run on any cloud or data center. However, each cloud has its own proprietary APIs to manage resources and deploy applications. For example, the Amazon Web Services (AWS) API uses an Elastic Compute Cloud (EC2) instance (a server) when provisioning a server, while Google Cloud's API uses a Google Compute Engine (GCE) VM (a server). Even the names of the resources are different! This is where Kubernetes plays a key role.
The wide adoption of Kubernetes has forced major cloud vendors to come up with tight integration solutions with Kubernetes. This allows anyone to spin up a Kubernetes cluster in AWS, GCP, or Azure in a matter of minutes.
The Kubernetes API enables you to manage cloud resources. Using the standard Kubernetes API, you can deploy applications on any major cloud provider without needing to learn about the cloud provider&apos...