Your company relies on Google Kubernetes Engine (GKE) for managing its various workloads, which include batch, stateful, and stateless services. Currently, all these workloads run on a single large GKE cluster that comprises a node pool of 200 nodes. Due to financial constraints, your company is seeking ways to reduce the operational costs associated with this GKE cluster but aims to ensure that the availability of these services is not compromised. What should you do?