Redpanda 23.2 introduces a new feature called follower fetching, which allows consumers to fetch records from the closest physical replica of a topic partition, thereby optimizing performance and reducing cloud infrastructure costs. This feature is particularly beneficial for large streaming data clusters, as it minimizes cross-availability zone (AZ) network traffic, which can be costly on cloud platforms like AWS. By enabling follower fetching, companies can significantly cut down on network expenses, as illustrated by a hypothetical example of Acme Corporation, which experiences a 60% reduction in monthly networking costs, translating to a 53% annual saving in AWS infrastructure costs. Follower fetching allows consumers to read data from any replica in the cluster, optimizing network traffic and minimizing read latency. The feature is currently in beta and is expected to become generally available soon, offering companies substantial savings by configuring their Redpanda clusters and nodes to support rack awareness and client rack properties.