Abstract: Data is split into a set of data packets and transmitted between a client computer system and a network service via a packet-switched network. The client computer system identifies a role, permission, group, or other credential that is associated with the data packets, and attaches a credential identifier such as a digital signature to the packets before they are transmitted over the network. A network service receives the data packets, and is configured to filter or route the data packets to a recipient using the attached credential identifier. The network service can adjust the filtering or routing process to occur within a data link, network, transport, or application layer. In some examples, the filtering or routing is provided from within a hypervisor.