At it's heart, Kore is built to support distributed applications. This is reflected in it's architecture - think of every instance of Kore as a "node". It can connect to other nodes, or store local data, or act as a server to listen for incoming connections.
Kore manages the storage and synchronization of logs for your application.
As you can see, a Kore node can store logs locally, and/or can connect to other nodes to synchronize logs, and/or can listen for connections from other nodes.
To learn how to use Kore in all these modes, go through the "Getting Started" document.