Agent Transfer

When multiple agents co-exist on a device, customers perceive agents as separate from each other, each with their own capabilities, invocations, and experiences. Sometimes, though, customers might make a request to an agent that it cannot directly fulfill. One way to help the customer in that situation is to implement an interoperability pattern called Agent Transfer.

During an Agent Transfer, the customer makes a request of an agent (Agent 1) who cannot directly fulfill their request (e.g. “I can’t do that”). However, if Agent 1 is aware of another agent (Agent 2) on the device which can likely fulfill that request, Agent 1 can summon the other agent to assist the customer. No data or context is passed between agents during a transfer, and the customer repeats their request directly to Agent 2 without needing to say the wake word. This pattern reduces some of the friction customers may face in completing their tasks.


It is important to use appropriate landmarking and sound cues, as shown in the example below, to make it clear to the customer that a transfer is taking place.